dna 2825f - dtic00 dna 2825f q4 final report december 1971 correlation of radar echoes from the...
TRANSCRIPT
00 DNA 2825Fq4 Final Report December 1971
CORRELATION OF RADAR ECHOES FROMTHE AURORA WITH SATELLITE-MEASUREDAURORAL PARTICLE PRECIPITATION
By: W. G. CHESNUT J. C. HODGES R. L. LEADABRAND
Prepared for:
DEFENSE NUCLEAR AGENCYWASHINGTON. D.C. 20301
CONTRACT DASA01-67-C-0066
APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.
NATIONAL TECHNICALINFORMATION SERVICE
S-"AN Va 22153
STANFORD RESEARCH INSTITUTEMenlo Park, California 94025, U.S.A.
211
UNCLASSIFIED
DOCUMENT CONTROL DATA. R ________D___
Security *Iaassfacie,on of title. body' of Ab~terI .#eud I idexites! .mnnuggn ntust be .ftfr,. d wri 4, is,. t,.'.-r..U) rmp.rt ,. ti.,'.aed)
IORIGINA TING AC TIVITV (Corporatq uthor) 1 Z.C~4SECUF41TY CLASSIO-I'AT1,OI
Stanford Research Institute UnclassifiedMenlo Park,- California 94025 IZIC14U /
I REPORT TITLE
CORRELAT ION OF RADAR ECHOES FROM THE AURORA WITH SATELLITE-kEASURED AURORALN
PARTICLE PRECIPITATION4. DESCRIPTIVE NOTES (Tr'p of tepee: and inclusive dae&i)
Final Report Covering the period M~rch 1969 to December 1971L U THO4ORIS (First name. ahidcej injitia. last nameW
Walter G. Chesnut James C. Hodges Ray L. Leadabrand
6RPRDAE70. TOTAL NO OF PAGES Tb. tNo Or REFS
December 1971 121SOCONTRACT Oft GRANT NO *.OflIGI*4ATOORS REPORT Nu"BEIIIS)
Contract DASA-0l-67-C-0066 Final ReportNWET K11BAXH SRI Project 6548 :
- ~~~Task and Subtask X532 thais "oet rCS An fEtnmbr hIm) eesa
e. Work Order 01 DNA 2825F
Approved for public release; distribution unlimited.
III SUPPLEmEriTARY NOTES 12 SPOT.SORINGSMILE IARN AC TI VITY
Defense Nuclear Agency
jWashington., D.~C.**13 ABSTRACI
The characteristics of auroral particle prec~ipitation as measured during 14 passes bythe Lockheed OVl-18 satellite over Alaska have been compared writh. radar aurora simul-taneously detected at 139 and 398 Wiz beneath the satellite orbit. The radar aurorawere measured using the SRI auroral radar located at Homer in southern Alaska. Thesatellite measured precipi~ation of protons with energies above 10 keV and precipita-tion of electrons with energies above 800 eV. All measurements were made. nea.r localmidnight while the E.-region was in darkness.* The study has revealed the following:(1) Protons with energies above 10 keV regular~iy pr-ec-ipitate to the south of elec-
'rons with energies above 800 eV.(2) Electron precipitation in regions of radar aurora seems sufficiently intense toI
produce the E-region ionization needed to explain the observed radar scattering.I141*(3) Radar aurora (on 11 of 14 passes) was not associated with regions of intense pro-
ton precipitation.
(4) No obvious relationship existed between regions of radar atarora and regions ofmost intense flux of electron precipitation power (number f lax times nverage
electron energy), though there may be a texdency for these to be spatially
anticorrelated.(5) There seems to be a tendency for the elec-crons incideut into r'qgions of radar
aurora to have a slightly lower average energy (order of W80) than electronsprecipitaiitng elsewhere.
DD, "U6,47 E ASSIFIEDSIPE 0101-407.60O1 S9cuithvClassification
UNCLASSIFIEDSecurity Class-f;cation
I KEY WORDS A LINK, LINK C
ROLE *l ROLE WT ROLE WT
Aurora
Aurora correlation
Auroral radar
Satellite-measured auroral precipitation
V I
[I
t 4i
DD 1473 (BACK) -
(PAGE 2) securiky Classification
.:I .1-i. .. , _-
Item 13 ABSTRACT (continued)
(6) There is a tendency for the average electron precipitation power to
be less in regio-: of radar aurora than in other precipitation
There does not seem to be a deterministic relationship between the char-
acter of precipitating electrons and radar aurora--only subtle, statistical
tendenc.es were found.
I
Il
. I
Final Report DNA 2825FDecember 1971
'4
CORRELATION OF RADAR ECHOES FROMTHE AURORA WITH SATELLITE-MEASUREDAURORAL PARTICLE PRECIPITATION
By: W. G. CHESNUT J. C. HODGES R. L. LEADABRAND
Prepared for:
DEFENSE NUCLEAR AGENCYWASHINGTON, D.C. 20301
CONTRACT DASAOI-67--C-0066
This work sponsored by the Defense Nuclear Agency under NWET Subtask K11BAXHX532.
SRI Project 6548
Approved by: 45
DAVID A. JOHNSON. Directo:-
Radio Physics Laboratory
RAY L. LEADABRAND. Executive Director
Eiec.'vnics and Radio Scienae Divaion
CopY No ......... .
I.
ABSTRACT
The characteristics of auroral particle precipitation as measured
during 14 passes by the Lockheed OVl-18 satellite over Alaska have been4
compared with radar aurora simultaneously detected at 139 and 398 MHz
beneath the satellite orbit. The radar aurora were measured using the
SRI auroral radar located at Homer in southern Alaska. The satellite
measured precipitation of protons with energies above 10 keV and precipi-
tation of electrons with energies above 800 eV. All measurements were
made near local midnight while the E-region uas in darkness. The study
has revealed the fcllowing:
(1) Protons with energies above 10 keV regularly precipitate
to the south of electrons with energies above 800 eV.
(2) Electron precipitation in regions of radar aurora seems
sufficiently intense to produce the E-region ionization
needed to explain the observed radar scattering.
(3) Radar aurora (on II of 14 passes) was not associated with
regions of intense proton precipitation.
(4) No obvious relationship existed between regions of radar
aurora and regions of most intense flux of electron
precipitation power (number flux times average electron
energy), though there may be a tendency for these to be
spatially anticorrelated.
(5) There seems to be a tendency for the electrons incident
into regions of radar aurora to have a slightly lower
average energy (order of 89M) than electrons precipitating
elsewhere.
iii
(6) There is a tendency for the average electron precipitation
power to be less in regions of radar aurora than in other
precipitation regions.
There does not seem to be a deterministic relationship between the
character of precipitating electrons and radar aurora--only subtle,statistical tendencies were found.
iv
CONTENTS
ABSTRACT .. . . .. .. ... .. ...... ....... . . iii
LIST OF ILLUSTRATIONS ............ ..................... .... vii
LIST OF TABLES ............... ......................... ... xvii
AChNOULEDGMENTS .................... ......................... xix
1. INTRODUCTION ................... ..................... 1
2. EXPERIMENT GEOMETRY .............. .................. 5
2.1 Geometric Constraints on Radar Data ..... ....... 5
2.2 Geometric Constraints on Satellite Data .... ..... 6
2.3 Simultaneity of Radar and Satellite Measurements. 7
3. EQUIPMENT, OPERATION, AND SCHEDULE ............ ... 13
3.1 Homer Radar ....... ................... ... 13
3.2 OVl-18 Satellite ..... ................. .... 13
3.3 Operation and Schedule .................. ... 16
4. SATELLITE-RADAR-DATA COMPILATION ....... ........... 23
4.1 Maps of Alaska ...... .................. ... 24
4.2 Plots of Radar-Aurora Signal-Intensity .......... 26
4.3 Satellite Data ...... .................. ... 27
4.3.1 Electron-Precipitation Data ..... ........ 27
4.3.2 Proton-Precipitation Data ........ ......... 2S
4.3.3 Histograms of Electron-Precipitation
Data ........... ................ ... 31
4.4 Comments on Specific Satellite Passes ...... .. 32
4.4.1 S.tellite Pass A388 ............... ... 32
4.4.2 Satellite Pass A431 ............... .... 33
4.4.3 Satellite Passes A2979 and A2995 ..... .. 34
VA
I_ _ _ i
5. COMPARISONS OF RADAR AURORA WITH MAGNOMETER AND
ALL-SKY-CAMbERA DATA. .. .. ... .......................... 1075.1 Radar-Aurora Magnetomoter Comparison ..... ...... 107
5.2 Rada,-.Aurara-All-Sky-Camera Comparison .... .... 108
6. S• ".Z OF SATELLITE DATA ..... ............... ... 113
S6.1 E-Region Ionization Produced by PrecipitatingElectrons .......... .................... ... 113
6.2 L-Region Ionization Produced by Precipitating
Protons .......... .................... ... 116
6,3 CVrrelation of Radar Echoes with Satellite-Borne
Particle-Precipitation Measurements ....... ... 1186.3.1 Corrilation of Regions of Peak Energy
Flux with Locations Rada'." Aurora ...... 118
6.3.2 Preouction of Nighttime E-Region Ioniza-tion by Particle Precipitation ...... ... 125
6.4 Discussion of Correlation Study and OtherPossible Relationships. .............. 127
7. CONCLU3IONS. .......... ...................... .... 137
8. SUGGESTIONS FOR FURTheR WORK ............. 139
REFERENCES ............ ........................... ...... 141
DISTRIBUTION LIST ......... ........................ ... 143
DD Ferm 1473
vi
ILLUSTRAT IONS
Figure 2.1 Map of Alaska Showing Magnetic-Aspect Angles for
the Homer Radar Magnetic L Shells ...... ......... 8
Figure 2.2 Elevation View of Magnetic Geometry in Plane Con-
taining Homer and College, Alaska ...... ......... 9
Figure 2.3 Map of Alaska Showing Overlap of Radar and All-
Sky-Camera Coverage ..... ................ ... 10
Figure 2.4 Profile View Showing How Satellite Data and Radar
Aurora Data are Projected Along Magnetic Field
Lines to Ground Level for Purposes of Comparison . 11
Figure 3.1 Photograph of Humer Radar ............... .... 14
IPFigure 3.2 Block Diagram of 139-MHz and 398-MHz Radars .... 21
Figure 4.A.1 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral DistributionsDuring Pass A518 ....... ................. ... 35
Figure 4.A.2 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral DistributionsDuring Pass A641 ..... ................. ... 36
Figure 4.A.3 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral Distributions
During Pass A624 ..... ................. ... 37
Figure 4.A.4 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Fnergy Flux
Along the Orbit, and Radar Auroral Distributions
During Pass A157 ....... ................. ... 38
vii
j
Figure 4.A.5 Map of Alaska Showing Ground Projection of
Satellite Orbit, Precipitating Electron EnergyFlux Along tne Orbit, and Radar Auroral Distri-
butions During Pass A388 ................ .... 39
Figure 4.A.6 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral Distributions
During Pass A232..... .................... 40
Figure 4.A.7 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral DistributionsDuring Pass A676 ..... ................. .... 41
Figure 4.A.8 Map of Alaska Showing Ground Pro.)ection of Satel-lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral DistributionsDuring Pass A751 ..... ................. .... 42
Figure 4.A.9 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy FluxAlong the Orbit, and Radar Auroral Distributions
During Pass A369 ...... ................ ... 43
Figure 4.A.10 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral Distributions
During Pass A657 ..... ................. .... 44
Figure 4.A.il Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit. and Radar Auroral Distributions
During Pass A431 ...... ................ ... 45
Figure 4.A.12 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral Distributions
During Pass A2964 ..... ................ ... 46
Figure 4.A.13 Map of Alaska Showing Ground Projection of Satel-lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral Distributions
During Pass A2995 ..... ................ ... 47
~31 viii
Figure 4.A.14 Map of Alaska Showing Ground Projection of Satel-
lite Orbit, Precipitating Electron Energy Flux
Along the Orbit, and Radar Auroral Distributions
During Pass A2979 ....... ................ ... 48
Figure 4.B.0 Calibration of Radar A-Scope Traces for Use with
Figures 4.B.1 through 4.B.14 ............. .... 49
Figure 4.B.1 Radar A-Scope Records for Satellite Prss A518 . 50
Figure 4.B.2 Radar A-Scope Records for Satellite Pass A641 . . 51
Figure 4.B.3 Radar A-Scope Records for Satellite Pass A624 . . 52
Figure 4.B.4 Radar A-Scope Records for Satellite Pass A157 . . 53
Figure 4.B.5 Radar A-Scope Records for Satellite Pass A388 . 54
Figure 4.B.6 Radar A-Scope Records for Satellite Pass A232 . . 55
Figure 4.B.7 Radar A-Scope Records for Satellite Pass A676 56
Figure 4.B.8 Radar A-Scope Records for Stellite Pass A751 .. 57
Figure 4.B.9 Radar A-Scope Records for Satellite Pass A369 . . 58
Figure 4.B.10 Radar A-Scope Records for Satellite Pass A657 . . 59
Figure 4.B.11 Radar A-Scope Records for Satellite Pass A431 . . 60
Figure 4.B.12 Radar A-Scope Records for Satellite Pass A2964. . 61
Figure 4.B.13 Radar A-Scope Records for Satellite Pass A2995. . 62
Figure 4.B.14 Radar A-Scope Records for Satellite Pass A2979. . 63
Figure 4.C.1 Precipitating Electron Number Flux (TSUM),
Average Energy (SI), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A518 . . 64
Figure 4.C.2 Precipitating Electron Number Flux (TSUM),
Average Energy (W3N), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A641 . 65
ix
Figure 4.C.3 Precipitating Electron Number Flux (TSUM),
Average Energy (EMN), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A624 . . 66
Figure 4.C.4 Precipitating Electron Number Flux (TSUMI),
Average Energy (EMN), and Energy Flux (ELE)
Versus Universal Time for Satellite Pass A157 . . 67
Figure 4.C.5 Precipitating Electron Number Flux (TSUI),
Average Energy (EDN), and Energy Flux (ELE) Ver-
sus Universal Time for Satellite Pass A388. . .. 68
Figure 4.C.6 Precipitating Electron Number Flux (TSUBM),Average Energy (EMN), and 1ncrgy Flux (ELE)
Versus Universal Time ior Satellite Pass A232 . . 69
Figure 4.C.7 Precipitating Electron Number Flux (TSUM),Average Energy (EMN), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A676 .. 70
Figure 4.C.8 Precipitating Electroa Number Flux (TSUM),Average Energy (EMN), and Energy Flux (ELE)
Versus Universal Time for Satellite Pass A751 . 71
Figure 4.C.9 Precipitating Electron Number Flux (TSUM),
Average Energy (EBN), and Energy Flux (ELE)
Versus Universal Time for Satellite Pass A369 72
Figure 4.C.10 Precipitating Electron Number Flux (TSUM),
Average Energy (EMN), and Energy Flux (ELE)
Versus Universal Time for Satellite Pass A657 . 73
Figure 4.C.11 Precipitating Electron Number Flux (TSUM),Average Energy (EM), and Energy Flux (ELE)Versus Universal Time for Satellite Pass A431 . 74
Figure 4.C.12 Precipitating Electron Number Flux (TSUM),Average Energy (MIN), and Energy Flux (ELE)
Versus Universal Time for Satellite Pass A2964. 75
Figure 4.C.13 Precipitating Electron Number Flux (TSIMhI),
Average Energy (1201), and Energy Flux (EJS)Versus Universal Time for Satellite Pass A2995. 76
A
Figure 4.C.14 Precipitating Electron Number Flux (TSUM),
Average Energy (EMN), and Energy Flux (ELE)
Versus Universal Time for Satellite Pass A2979. 77
Figure 4.D.1 Precipitating Proton Number Flux (B), Average
Energy (M), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV) vs. Universai Time for
Satellite Pass A518 ........ ............... 78
Figure 4.D.2 Precipitating Proton Number Flux (B), Average
Energy (E), Statistical Error in Average Energy
(BE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV) vs. Universal Time for
Satellite Pass A641 .... ........... ...... 79
Figure 4.D.3 Precipitating Proton Number Flux (B), Average
Energy (1), Statistical Error in Average Energy
I (AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV) vs. Universal Time for
Satellite Pass A624 ..... ............... ... 80
Figure 4.D.4 Precipitating Proton Number Flux (B), Average
Energy (E), Statistical Error in Average Energy
(AE). Energy Flux (PAV), and Statistical Error
in Energy Flux (6PAV) vs. Universal Time for A
Satellite Pass A157 ..... ............... ... 81
Figure 4.D.5 Precipitating Proton Number Flux (B), Average
Energy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV) vs. Universal Time forSatellite Pass A388 ..... ............... ... 82
SFigure 4.D.6 Precipitating Proton Number Flux (B), AverageEnergy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APWV) vs. Universal Time forSatellite Pass A232 ...... .............. ... 83
Figure 4.D.7 Precipitating Proton Number Flux (B), AverageEnergy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV), vs. Universal Time forSSatellite Pass A676 ...... .............. . .. 84
I xi
MI
IN
Figure 4.D.8 Precipitating Proton Number Flux (B), Average
Energy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV), vs. Universal Time for
Satellite Pass A751 .... ............... ... 85
Figure 4.D.9 Precipitating Proton.Number Flux (B), Average
Energy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV), vs. Universal Time for
Satellite Pass A369 ...... .......... .. ... 86
Figure 4.D.10 Precipitating Proton Number Flux (B), Average
Energy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV), vs. Universal Time for
Satellite Pass A657 ..... .............. ... 87
Figure 4.D.11 Precipitating Proton Number Flux (B), AverageEnergy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Error
in Energy Flux (APAV), vs. Universal Time for
Satellite Pass A431 .... ............... ... 88
Figure 4.D.12 Precipitating Proton Number Flux (B), Average
Energy (E), Statistical Error in Average Energy(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APAV), vs. Universal Time for
Satellite Pass A2964 .... ............... .... 89
Figure 4.D.13 Precipitating Proton Number Flux (B), Average
Energy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APAV), vs. Universal Time for
Satellite Pass A2995 .... ............... .... 90
Figure 4.D.14 Precipitating Proton Number Flux (B). Average
Energy (E), Statistical Error in Average Energy
(AE), Energy Flux (PAV), and Statistical Errorin Energy Flux (APAV), vs. Universal Time for
Satellite Pass A2979 ..... ............... .... 91
Figure 4.E.1 Histograms of Logaritiniic Occurrence vs. Precipi-
tating Electron Number Flux (TSUM), Energy Flux(ELE), and Average Energy (EMN) for Satellite
Pass A518 .......... ................... ... 92
xii
I
Figure 4.E.2 Histograms of Logarithmic Occurrence vs. Precipi-
tating Electron Number Flux (TSUM), Energy Flux
(ELX), and Average Energy (DIN) for Satellite
Pass A641 ........ .................... ... 93
Figure 4.E.3 Histograms of Logarithmic Occurrence vs. Precipi-
tating Electron Number Flux (TS'JI), Energy Flux
(ELE), and Average Energy (EMN) for Satellite
Pass A624 ........ .................... ... 94
Figure 4.E.4 Histograms of Logarithmic Occurrence vs. Precipi- Itating Electron Number Flux (TSUM), Energy Flux
(ELE). and Average Energy (EIM) for Satellite
Pass A157 ........ ................... ... 95
Figure 4.E.5 Histograms of Logarithmic Occurrence vs. Precipi-
tating Electron Number Flux (TSUM), Energy Flux
(ELE), and Average Energy (EWIN) for Satellite
Pass A388 ....... ................... .... 96
Figure 4.E.6 Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSUIM), Energy Flux
S(EIE), and Average Energy (MIN) for Satellite
Pass A232 . ................................... 97
Figure 4.E.7 Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSLTAI), Energy Flux
(ELE), and Average Energy (EIN) for SatellitePass A676 ...... ................. ..... 98
Figure 4.E.8 Histograms of Logarithmic Occurrence vs. Precipi- Atating Electron Number Flux (TSUM), Energy Flux
(ELE), and Average Energy (BN) for Satellite
Pass A751 ........ .................... ... 99
Figure 4.E.9 Histograms of Logarithmic Occurrence vs. Precipi-
tating Electron Number Flux (TSUM), Energy Flux(ELE), and Average Energy (EMN) for Satellite
Pass A369 ........ ................... .... 100
Figure 4.E.1O Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSVID, Energy Flux
(ELE), and Average Energy (MIN) for SatelliteSPass A657 . . ..... .................. .. 101
xiii
13
Figure 4.E.11 Histograms of Logarithmic Occurrence vs. Precipi-
tating Electron Number Flux (TSUM). Energy Flux
(ELE), and Average Energy (EMN) for Satellite
Pass A431 .. .. ....................................102
Figure 4.E.12 Histograms of Logarithmic Occurrence vs. Precipi-tating Electron Number Flux (TSUM), Energy Flux
(ELE). and Average Energy (EMN) for Satellite
Pass A2964 ....... .................... .... 103
Figure 4.E.13 Histograms of Logarithmic Occurrence vs. Precipi-
tating Electron Number Flux (TSUM), Energy Flux
(ELE), and Average Energy (EMN) for Satellite
Pass A2995 ....... .................... .... 104
Figure 4.E.14 Histograms of Logarithmic Occurrence vs. Precipi-
tating Electron Number Flux (TSUM), Energy Flux
(ELE), and Average Energy (MI) for SatellitePass A2979 ....... ....................... 105
Figure 5.1 Simultaneous Auroral Radar Echo Strength and
College, Alaska Magnetometer Deflections v.: Time
on 24 March 1969 ....... ............... . ... O110
Figure 5.2 Map Shouing Location of Visual Aurora and Radar
Aurora at 12:19 UT. 24 September 1969. Radarfrequency, 139 MHz ......... ............... 11
Figure 5.3 Map Shoulng Location of Visual Aurora and Radar
Aurora at 12:28 UT, 10 October 1969. Radar fre-
quency, 139 MHz ............ ................. 112
Figure 6.1 Altitude of the Maximum Rate of Production of
Ionization vs. Monoenergetic Electron Energy for
Monoenergetic Electron Spectrum. isotropic
angular distribution over dowmward hemisphere . 130
Figure 6.2 Altitude vs. Electron Density for Various Mono-
energetic Electron Fluxes. Electron density is
at ,ktitude of peak ionization rate ....... ... 131
Figure 6.3 Altitude vs. Chemical Time Constant for Mono-energetic Electron Fluxes. Time constant is
appropriate to altitude of peak ionization for
mongenergetic electrons ..... ............ ... 132
xiv
3_
- - -- ~-~-~---~~-- -zA
Figure 6.4 Altitude of Maximum Rate of Production of Ioniza-
tion vs. Mean Proton Energy for an Exponential IProton Spectrum, Isotropic Angular Distributionover Doumward Hemisphere .... ............ ..... 133
Figure 6.5 Electron Density at 109 km Altitude vs. Proton
Flux. Proton spectrum--exponential with E = 32
keV. Pitch-angle distribution is isotropic over
downward hemisphere ...... ................ ... 134
Figure 6.6 Average Properties of Precipitating Electrons vs.Satellite Pass Sequence Number Arranged According
to Ascending Three-Hour Magnetic K-Index .... ...... 135
Figure 6.7 Ratio of the Average Properties of Precipitating
Electrons in Regions where Radar Aurora wasObserve1', to those Properties in Regions where No
Radar Aurora was Observed, vs. Satellite-Pass
Sequence Number Arranged According to AscendingThree-Hour Magnetic K-Index ... ........... .... 136
ININ
xv
TABLES
Table 3.1 Homer Auroral Radar Parameters ... ........... ... 14
Table 3.2 Summpry of Characteristics of Satellite Passes
Used in this Study ....... ................ ... 17
Table 3.3 Simultaneity of Data Periods for Satellite Pass
and Radar Data ....... .................. ... 18
Table 4.1 Summary of Satellite Passes Ordered According to I
Magnetic Activity ....... .................... 25
Table 4.2 Summary of Definitions of Various Satellite Data
Quantities ....... ..................... .... 28
Table 6.1 Coincidences sod Anti-Coincidences Between Radar
Aurora and Peaks in Energy Precipitation from Satel-
lite Measurements ............ .................. 119
Table 6.2 Coincidences and Anti-Coincidences Between Radar
Aurora and Measurable Particle Precipitation 126 12
xvii
i7I
ACKNOWLEDGMENTS
We wish to thank Mr. Ronald Presnell of our Laboratory for the
assistance provided to this program in a number of ways. In particular,
Mr. Presnell has continued to modernize the Homer radar system and has
provided guidance on the operation of the instrument.
Miss Judy Moore provided computer programming necessary for the
reduction and presentation of the satellite data as it is found in this
report.
Dr. Richard Sharp and Dr. Richard Johnson of the Lockheed Space
Sciences Laboratory of Palo Alto, California, furnished the satellite
electron and proton precipitation data. We appreciate their guidance
and suggestions concerning the use of their data.
We appreciate the continued guidance and interest shown by Mr. Dow
Eelyn of the Defense Nuclear Agency.
The Geophysical Institute, College, Alaska, furnished all-sky
photographs.' The U.S. Coast and Geodetic Survey, College, Alaska,
furnished magnetograms.
The 1969 Radar Data Collection Program was supported both by the
Defense Nuclear Agency and by the National Science Foundation (Grant
GA-1061).
xix
I ___________ _____
1. INTRODUCTION
This report describes the research performed by the Rado Physics
Laboratory of Stanford Research Institute for the Defense Nuclear Agency
on Contract DASA01-67-C-0066. The program compares Alaskan auroral radar
echoes with the OVI-18 satellite data gathered by Lockheed Space Sciences
Laboratory.
The experiment is referred to as an input-output experiment. The
input to the ionosphere is measured by the satellite, and the resulting
effects (the output) are simultaneously measured by the Homer radar.
The program is a continuation of work started in 1965 under DNA sponsor-
ship. 1 The earlier experiment produced only one pass during which radar
aurora were present before the satellite orbit decayed.
S~The data used in this report were obtained under DNA sponsorship
between 24 March 1969 and 8 May 1969, and under NSF sponsorship from 3
September to 17 October 1969. These periods were near the maximum of the
11-year solar sunspot activity cycle. The data from 14 of approximately
60 satellite passes were made available for SRI by Lockheed late in March
& 1971. Five hours of magnetic data supplied by the Coast and Geodetic
Survey, College, Alaska, and five days of all-sky-camera data furnished
by the Geophysical institute of the University of Alaska are correlated
wi•th radar data in this report.
Although radar echoes have been observed from the aurora for approxi-
imately 20 years the mechanism that produces these radar echoes is stillRA
=*References are listed at the end of the report. S
ILI R
_ _ W:
It
not well understood. Earlier studies of radar data correlation with other
radio data, visual data, magnetic-field-fluctuation data, and drta from
other geophysical phenomena have not always shoun clear-cut relationships.
In particular, comparisons of visual aurora and radar aurora have showm
that the radar aurora is usually mucb broader geographically and usranly
occupies a somewhat different position in space than the visual aurora.•
Correlating satellite-measured particle precipitation with radar
aurora seems to be a fruitful technique to better investigate the under-
lying physics. The OVl-18 satellite that was used in this experiment
measured electron and proton precipitation at an altitude of 500 km over
Alaska. The Homer radar collected data magnetically below the satellite
during the time of the pass.
A report comparing precipitation during one daytime pass with radar
cchoes obtained at 1300 MHz by the Miillstone Hill radar has been reported
by Hagfors et al. 3 Their comparison also used the OVI-18 satellite. For
this one daytime pass they find that the radar aurora wa:; located in :,
region of proton precipitation. We find, and shall show later, that on
the night side radar aurora and proton precipitation tend to taLe place
at different latitudes.
This report is organized as follows. Section 2 discusses the experi-
ment geometry. The geometrical constraints on collection of radar data
and the geometric considerations relevant to radar/satellite comparisons
are presented.
Section 3 describes the characteristics of the Homer radar in suffi-
cient detail to understand this report. The characteristics of the
counters that were carried on the OVl-18 satellite are also described.
Section 3 also describes the operation of the equipment and the data-
collecting schedule.
2
i!j
Section 4 is a Cata compilation of satellite and radar data for 14
satellite past3s for which satellite particle-counter measurements were
provide%: to SRI by the Lockheed Corporation. The correlations that are
made have to do with the spatial location of radar and precipitating
particles.
Section 5 provides a comparison of some of our radar data with mag-
netometer measurements msde underneath the radar auroral scattering.I Section 5 alsc presents a comparison of the geometry of visual auroral
arcs with radar aurora obtained at the same time.
Section 6 is a study of the satellite data and correlations made
between the satellite and radar data in an atte:mpt to understand the nature
of the phenomena that are being observed. In particular, we concern
ourselves with production of ionization in the nighttime E-region, since
this ionization is necessary for the existence of radar scattering. We
also concern ourselves with the location of radar aurora and the character
of the particle precipitation.
Section 7 presents the conclusions that have been reached as a re-I
sult of this research effort.
Section 8 indicates additional work that should be done in order to 1
better understand the data that have already been collected.
3
IT
I
2. EXPERIMENT GEOMETRY
!IA variety of geometric considerations affect the interpretation of
the data correlations covered in this report. In this section we describe
some of these constraints so that the reader can view our data in better
perspective.
2.1 Geometry Constraints on Radar Data
As is well known., radar aurora is only obtained when the radar
beam line is nearly perpendicular to the earth's magnetic field in the
scattering volume. The signal strength depends on the magnetic aspect
angle; this angle is the difference between the angle exactly perpendicular
to the magnetic field and the actual angle between the radar beam line.
The Homer radar was located in order that this magnetic geometry constraint
could permit as wide a coverage of radar aurora over the state of Alaska
as possible. Figure 2.1 presents a plot of the magnetic aspect angle Tor
aurora scattering at en altitude of 110 km as observed from the Homer radar
site. Homer is located near the bottom of the map. As a practical matter,
radar auroras are usually detectable only When they occur within the 30
contour shown in Figure 2.1.
Figure 2.2 presents an elevation view. which is another way to
demonstrate the magnetic geometry. This view shows that the 00 magnetic -
aspect angle is a function of altitude when the radar views directly North.
Figure 2.3 presents a map of Alaska showing the 30 magnetic
ra aspect angle as seen from the Homer radar site for scattering at 110 kLmI
Figures are grouped at the end of each major section.
I'KLuk.,~jAf'PAGE BLANK 5ru
altitude. Also shown on the map is the area covered by an all-sky cameca
located at Fairbanks. The figure shows overlapping coverage. Another
all-sky camera, located et Ft. Yukon. extends photographic coverage
another 200 km north.
Figure 2.3 presents in R qualitative way the area of Alaska and
Canada that can be surveyed by the Homer radar in a practical sense.
During very strong radar auroras, echoes can be obtaii.ed from outside
this contour, but such reception is rare.
2.2 Geometric Constraints on Satellite Data
Figure 2.4 shows the way in which satellite precipitation data
were geometrically related to the radar data. This figure represents
an elevation view along a meridian through Homer and College, Alaska.
The figure shows various magnetic field lines. We show schematically how
satellite measurements of precipitation are projected down magnetic field
lines to the earth's surface. The radar aurora are also projected down
magnetic field lines to the earth's surface. Therefore, comparisons that
will be seen later will be between precipitation and radar aurora as both
are projected to the earth's surface.
There are three other geometry considerations concerning satel-
lite data that must be borne in mind, First, we project electron-
precipitation effects to the earth's surface as described by Figure 2.4.
!.- fact, ionization will be produced along magnetic field lines where the
electrons are precipitating. Therefore, there will be a real, geometrical
relationship in the radar scattering region, between electrons and radar
aurora. This is not the case for protons. Satellite-measured proton
precipitation is made at altitudes above 400 kilometers. Protons that
precipitate to lower altitudes where the atmosphere is more dense .. ill
suffer charge exchange. As a result, a very localized proton fluz at
6
satellite altitude nay be spatially spread by several hundred kilometers
by the time the protons reach altitudes between 100 and 11' kilometers.
SITherefore, relationships between regious of peak proton precipitation,
Sas measured by the satellite, and regions of radar aurora do not neces-
sarily relate to overlap of E-region ionization produced by the protons.
The relationship, if one does exist, is then between the mechanism that
causes proton dumping and that which causes spatial structuring of E-
region ionization that leads to radar aurora.
- There is a third geometry consideration regarding coi relations
I between radar aurora and precipitation. This relationship has to do
with the production of the E-region ionization that is necessary for
radar aurora. Since radar aurora most often occurs at an altitude on
the order of 110 kilometers. precipitating particles must have sufficiently
high energy to penetrate to this altitude if they are going to contribute
to the E-region ionization.
2.3 Simultaneity of Radar and Satellite Measurements
When we perform correlation studies between radar aurora and
satellite precipitation, it is essential, of zourse, that the measure-
ments be made at the same time. Since radar aurora can move large dis-
tances in a time of several tens of seconds, one surmises that the
simultaneity accuracy ought to be within 10 or 20 seconds. Pre-pass
ephemeris data were used in our analyses. Pre-pass ephemeris data was
sometimes in error. Four cases were particularly poor ia their simul-
_ •taneity. These four cases were passes A388, A431, A513, and A641. The
simultaneity for the last two cases uas in error by 7 and 15 minutes,
I respectively.
j Satellite precipitation data are output and recorded once per
second. During this time period the satellite moves about 7-1/2 km.
1 7
s1Satellite data are referenced to "satellite tir-r;." The post- -
pass satellite ephemeris is also related to this time. It is therefore
very important that the ephemeris geographical error be less than this;
otherwise our radar/satellite comparisons will not spatially overlap
properly. We assume Lockheed data are this accurate.
eoow
Lefr "L-7 1
LSHKELLSh - O0km .
S~.'• OFF I. ONTOURS.4h- 110 krn•;
AZIMUTH* L4STANUM PRIOJECTMO
FROM HOW"E
i FIGURE 2.1 MAP OF ALASKA SHOW'NG MAGNETIC-ASPECT ANGLESS~FOR THE HOMER RADAR MAGNETIC L SHELLS
I °aa
Aw
ii--iI
S3. EQUIPMENI\, OPERATION, AND SCHEDULE
In this section we give pErtinent details of the instrumentation
Sthat was used on this program. \We also indicate the data-taking schedule
for the satellite passes for whih Lockheed has provided satellite data
for our use.
3.1 Homer Radar
The Homer Radar is located on Ohlson Mountain about 13 miles
from Homer, Alaska. Figure 3.1 is a photograph of the radar complex.
The latitude and longitude of the radar are as follows: latitude 59.7142*N;
longitude 151.5333 0 W. Table 3.1 provides a brief summary of the param-
eters of this radar. The six radar frequencies shoun in the table are
fed through the common 60-foot-diameter paraboloid. During satellite
passes all radar frequencies were turned on, were operating, and data
were recorded. In this work we have utilized only the data obtained at
139 and 398 MHz. A far more detailed distribution of the radar may be
obtained from a number of references. 2 Figure 3.2 provides a block
diagram of the 139- and 398-MHz radar systems as they were configured
for these experiments.
3.2 0Vl-18 SatelliteI
-he OVI-18 satellite was instrumented by the Space Sciences
Laboratory of the Lockheed Missile System Corporation. Details of the
satellite may be found in Lockheed reports.",s The satellite was launched
into nevr polar orbit. The satellite gathered 48 channels of data. The
data obtained by Lockheed were corrected for a variety of factors before
I
PREC0!Nl PAGE BLANK 13
If
II0 - 00 0 1 I
C0 0 0 t- cs0 .m $ ito 1-4 CMJ l
03 .4
gp t!. .0m~~ ~ 0 m C3 o m
5) I 4 C ) -4 C 4)
.44I)3ASI. 04
0)03 0)54) 0 5" 41 0 if)C
54)3 -44
.0 .
CiC3
*0 m 4 C
a 4J~ 4J. 4J a 0 -0 0 ' 44 ) 0 0. 0
C4) A. ) P. 54 <q ~ ~ .
15.
-~ -- - -~ -- -
being sent to SRI. For instance, counting rates were converted to ab-
solute particle flux; data were selected from many sets of counters accord-
ing to the momentary satellite orientation (the satellite stabilization
mechanisms failed to operate correctly); average electron energy was
computed, etc.
Basically, the satellite recorded electron-precipitation flux
over a range from 0.8 to 37.0 keV. These data were obtained by five
channeltron electron counters, each of which selected electrons over a
narrow energy range. Due to peculiarities of the satellite system, some-
times data were supplied from only the four lower-energy channeltron
counters. These data covered the energy range from 0.8 to 16.3 keV. The
electron-precipitation data were provided to SRI as TSUM or PSUM, which
means the total electron flux in electrons per square centimeter per
second per steradian. The TSUM included all 5 channels; the PSUM included
the flux in the lower four energy channels. The electron-precipitation
data supplied to SRI also contained the average electron energy obtained
by computing the first moment of the energy spectrum; this is obtained
by multiplying the energy of each counter times its particle flux in the
five (or four) channeltron counters. This quantity we label EXN. The
meaningful counting threshold for these counters was about 5 X 106 counts
per second.
The proton-flux data that were transmitted to SRI by Lockheed
gave the absolute proton fluxes as obtained by two proton counters. The
first is labeled the B counter. The B-type counters measure absolute
proton flux for protons with energies above approximately 10 keV. The
data provided to ShI were given in protons per centimeter per second per
5steradian. Meaningful counting-rate threshold was between 1 X 10 and
4 x 10 per second, depending on which of the B counters was being used.
Another type of proto,. counter was labeled the D counter. These proton
counters measured absolute flux of protons with energies above 38 keV.
16
The flux provided SRI was in units of protons per square centimeter per
steradian per second. Meaningful counting-rate thresholds were I x 105
to 4 X 105 counts per second, again depending on which counter was being
used.
A table presented in the data compilation of Section 4 provides
a summary of these counter characteristics, along with other data that
are pertinent to understanding the data presentations given there.
3.3 Operation and Schedule
The usual operational procedure was to receive satellite emph,!m-
eris data froin Lockheed for expected satellite passes. These ephemeris
data give predicted satellite position versus time. Accordingly, the
Homer radar would provide a sweep of the appropriate regions of the sky
during the satellite pass. Before and after the satellite pass the radar
would survey the entire sky over Alaska in order to observe the larger
picture of overall activity.
There were many occasions during which radar and satellite data
were obtained simultaneously. For reasons beyond our control, Lockheed
has provided data from only 14 of these satellite passes. These 14 passes
were selected because radar aurora or satellite precipitation was measur-
able. Table 3.2 presents a detailed summary of the 14 passes.
Radar data were obtained over a period of time, from considerably
before the expected satellite pass (usually several hours) to a time very
late after the pass was ex:pected to be completed (again, ui-ually several
hours). The radar data analyzed in this report are for the time period
determined from the pre-pass ephemeris data. As a result, the simultaneity
of data comparisons between radar and satellite is not always very close.
Table 3.3 summarizes this simultaneity more concisely than is obvious
*Table 4.2.
17
0wC -05 1at n C - t
t~! - N i lO C.0 IO N in0 2. 2
-0 -0C 0
0. 40 a &.
E-0 t- 0, In. =' : - - N N
W a . 0 E0 I 3.T z. 0 -
4. ft 34 ft 0 .o c
.. %,. E F- as. k 0 ;2. a . 0 0 ti a
tc to Wa 0~ a a2 w - 2.o a a Im A
A C
0 0 ) v 3 - 1) 2
S. 00 1.0 l ca:W. w C' q 4.r , ) t
. .4C 0 cr m 0 to v- ol No m~ 0to n 0.20 02.- ý m 0 4m NI N1 M Iv v %. 0 a2v. v-. v v L, 0 L) in L,) L3 to rnin - S. -
C4 in N.%n ý -
- 4 0m u
0 4
a4 00 0 0 - 06 N
0S 00 2D m 0 Nl 0 q 1- m 0 in 5.
a coo c! 0T!0 N -La------------------------- -4 0- -)e, -3 c C4 0 .
0 N ainI0A 0 0 " :; ! -at l l
0- - - - - -M -- - - - - - - - - - -- M eq C4 Q
CL in N u n in - N i
0 0 04 N . Ci No V - n4c - n N - 0 - - 0 i mi n 0 U
ino a..................................m m.O 00 0 0 0 0 0 - NM
- - - -- - -
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _41
-W a ltA:K
IMI
cs~
C3r - C') 10 -I D -4 C' 3 C) 4 k --
'-0
0:
C3
C) 4.4 CS
IDA 0 Id. 0 0 0 0 0D Le) 0 ) in IDQ0 4O ca oa3~ C) C ) 4 0
03 <
44 0 4)
0 c
cs E- C
EC 4 C' 0 0 00 to 00
0 04.. .=) C:
41-
4 to 0t- I-4C) C40 003 - 0 IT %0 t.0 03 03 000 4 CS) 03 cm M IV 0 0 W 0 w 0 to W - N* N 3%
41
j 19
from Table 3.2. Simultaneity of passes A431, A518, and A641 was exceed-
ingly poor. For passes A518 and A641 the poor time simultaneity probably
does not matter since our experience shows that the low K indices imply, geographically stable phenomena. Pass A431 occurred during high activity,
so that lack of time simultaneity of the measurements makes our compari-
sons less valuable for this pass.
The data collection was near midnight local time in all cases
except in October 1969. During October 1969 coordinated measurements
were made on eight midday satellite passes. No radar aurora was observed
during any of these daytime passes. We do not know if the Lockheed
satellite recorded precipitation in these regions of the ionosphere during
these pnsses.
The radar data that were obtained were tape recorded on a 14-
channel analog instrument. The data were read back in analog fashion.
The radar data were measured, geometrically corrected, and plotted on
maps that will be presented later, in Section 4. The satellite data were
reduced by computer before Lockheed delivered the data to SRI. At SRI
a number of computations were made on the satellite data. The results
of these computations are presented in Section 4.
Z
20
- --- -~-- -
zL~ui
2 cc
I- co
wIw
CC 0 c0 x
w Cx 0 c
Lu L-j
0. -j-i up
0Ou w I )
U- ý- r-);o
cc~ Lu c -L 20
OD z tr21
coIto Zw tca 04
I It1 4. SATELLITE RADAR-DATA COMPILATION
Satellite precipitation data were provided to SRI for 14 nighttime
satellite passes during which measurable precipitation and radar aurora
were simultaneously present. Note that radar aurora was not present
when there was no precipitation. Also, recall that during October 1969
I coordinated measurements were made on eight midday satellite passes.
SDuring these daytime passes, no radar aurora was observed, though at this
writing we do not know whether the satellite observed any precipitation.
I• The data obtained are presented in this section in several ways.
The main effort is to provide information that allows for a comparative
study of the spatial locations of radar aurora and the simultaneously
Sobserved satellite precipitation measurements. There are five kinds ofjdata presentations. They will be described in more detail below. Briefly,
they are as follows:
* The first format is a map of Alaska with the projection of Ithe satellite orbit down the earth's magnetic field lines
onto the map. Also shown on the map are the areas in which
radar aurora was found. The maps give an indication of
where precipitation was occurring.
* The second data presentation is a series of A-scope photo-
graphs giving radar aurora intersity.
* The third and fourth formats provide various data concerning
electron precipitation and proton precipitation versus
orbital time.
PRECEDING PAGE BLANK23
IN
.The fifth format provides histograms of occurrence of
electron-number flux, electron-energy flux, and average
electron energy for each of the 14 passes according to
whether radar aurora was or was not present in the precipi-
tation region. For easy identification, the figures in
these five formats are designated by figure numbers containing
the letters A, B, C, D, and E, respectively.
All data are labeled according to satellite orbit number. Data are
presented in order of ascending three-houi magnetic index at College,
Alaska. Table 3.2 provides a schedule of satellite passes and other perti-
at information. The times shown on all figures are Universal Time
(Greenwich Meridian Time). Subtract 10 hours to convert to Alaska Standard
time. Thus it will be seen that all measurements reported here were made
between midnight and 2 AM local time in Alaska. Note that because the
Homer radar can observe aurora over a wide azimuth around magnetic north,
radar aurora may be measured from areas that are experiencing significantly
earlier or later local magnetic time tnan at the radar site itself.
Dr-in all measurements reported here the E-region was in darkness. Table
4. 1 sts satellite passes according to ascending K index. The table
summarizes other pertinent information.
4.1 Maps of Alaska
In Figures 4.A.1 through 4.A.14 we present 14 maps of Alaska,
one for each of 14 satellite passes. The maps show the location of the
Homer radar facility near -he bottom of each figure. Each map is labeled
with the date of the satellite pass, the satellite orbit number, and the
radar frequency for which clutter returns are shown. The satellite orbit
as projected down the earth's magnetic field lines to the earth's surface
is shown on the map. Tic marks indicate Universal Time along this orbi-
tal projection. The total energy precipitated as a function of orbital
24
I•
Table 4.1
SUWMARY OF SATELLITE PASSES ORDERED ACCORDINGTO MAGNETIC ACTIVITY
Subjective Character
Time Simultaneity of Precipitation
3-Hour of Radar and (See Section 6
Sequence Pass Magnetic Satellite Data for description)
No. No. Index-K (minutes:seconds) Electrons Protons
I A518 0 7:00 Poor ....
2 A641 0 14:25 Very poor -- none
3 A624 1 0:00 Good ....4 A157 1 0:15 Good ....
5 A388 1 2:20 Good E --
6 A232 2 0:30 Good E wp
7 A676 2 0:10 Good E --
8 A751 3 0:35 Good E --
9 A369 4 1:45 Fair E --
10 A657 5 0:15 Good E WP
11 A431 5 3:20 Poor E P
12 A2964 6 0:45 Fair E none
13 A2995 7 0:15 Good E P
14 A2979 7 0:10 Good E PII
position along the orbit is shown by means of cross-hatched are:... The
amplitude of the cross-hatching from the satellite orbit is linearly
proportional to energy deposition. In an effort to make the presentation
more legible, the cross-hatched areas are extended on both sides of the
orbit.
Radar aurora are shown on the map by speckled areas. The re-
gion from Which clutter was obtained, if it was well mapped by the radar,
is also outlined with a heavy line. In other cases data are presented
25
--- M
only along the satellite orbit. In these cases the radar aurora is shown
as speckled with heavy boundaries that are not closed.
On several satellite passes t',ere are special considerations.
These will be noted on the figuizq.
4.2 Radar-Aurora Signal-Intensity Plots
Radar auroral echo strength and location of radar echoes are
determined in this report from A-scope photographic records. An A-scopa
record presents signal intensity or signal voltage as a function of rauge
from the radar. The A-scope photographs presented are time-exposure
photographs of approximately 30 seconds duration. The figures are anno-
tated with radar frequency used, time duration of signal integration
(Universal Time--subtract 10 hours for Alaska Standard time.), and antenna
azimuth and elevation. The A-scope records pro-iide approximate logarithm
of signal amplitude (post-detection amplitude) versus range in kilometers.
On some figures we have placed the symbol ;, which indicates that during
the integration time the radar aurora amplitude seemed to vary in ampli-
tude in a significant way.
Figure 4.B.0 presents calibration signal amplitudes that may be
used with the A-scope photographs. The calibrations are given in terms
of power input into the receivers for both the 139- and 398-MHz radars.
Calibration must be done in this way rather than as radar cross section
because cross-section calibration would be a function of range, aurora
geometry, antenna beam pattern, etc. A value of -100 dBm corresponds
approximately to 1I m2/m . This relationship assumes radar aurora is
on the order of 10 kilometers in vertical thickness, at a range of about
800 kilometers, and is beam-filling in the azimuth direction. Figures
are labeled according to three-hour magnetic index in ascending order.
The radar signal amplitudes for 14 satellite passes are presented in
Figures 4.B.1 through 4.B.14.
26
_ _ _ I4.3 Satellite Data
The satellite data are presented in the form of computer-drawn
graphs. Data are presented for electron precipitation and proton precipi-
tation separately.
4.3.1 Electron-Precipitation Data
Electron-precipitation data versus orbital time (Univer-
sal Time--subtract 10 hours for Alaska Standard time) are presented in
Figures 4.C.1 through 4.C.14. In each figure are shown plots of three
separate quantities. The top part of each figure presents T-PSUM. The
ordinate in the top figure is logarithmic in electrons per square centi-
meter per second per steradian. In regions where no data points are
shoum, the counting rates were low; at the suggestion of Lockheed person-
nel we have imposed a threshold at a value near to or just equal to the
on-board, radioactive-calibration-source counting rate. In some cases,,
through oversight, the amount of time for which data were supplied to us
was less than we required.
The middle part of each figure presents as ordinate the
quantity MIN. This quantity is the mean energy of the precipitating
electrons as determined by the five (or four) energy-selecting electron
counters. The units are linear from 0 to 10 kV. The bottom part of each
figure -resents 3 its ordinate the logarithm of the product of the T-
PSUM times EMN. ,e units are electron-keV per square centimeter per
second per steradian. This curve provides a measure of energy flux (power)
as a function of position along the satellite orbit. We have indicated
on the figures, by cross-hatched regions along the trajectory, where
radar aurora simultaneously existed. Table 4.2 provides a summary de-
scription of the quantities given in Figures 4.C.1 through 4.C.14.
27
27
Table 4.2
SUMMARY OF DEFINITIONSOF VARIOUS SATELLITE DATA QUANTITIES
k IElectronsI TSUM Total electron flux between 0.8 and 37 keV as measured by
five energy-selecting channeltron counters. MeaningfulSi ~threshold: 5.106 counts/s. Units: el/cm2-s-ster.
PSUM Total electron flux between 0.8 and 16.3 keV as measured
by four energy-selecting channeltron counters. Meaningfulthreshold: 5.106 counts/s. Units: el/cm2 -
EMN Average electron energy determined from counting-rate dis-
tribution in five (four) channeltron counters.
ELE Total electron energy flux (power density): (TSUM X EMN,
or PSUM X EMN).
Protons
2_B Proton flux, protons/cm -s-ster. Approximately l0-keV
threshold, no upper limit. Meaningful counting-rate
threshold: 1 X 105 to 4 X 105 counts/s, depending on thecounter.
D Proton flux, protons/cm2 -s-ster, 38-keV threshold, noupper limit. Meaningful counting-rate threshold: 1 X 105to 4 X 105 counts/s, depending on the counter.
EAverage proton energy under the assumption that the proton
energy spectrum is exponential. This quantity is deter-
mined from the relative counting rates in the B and D
counters.
AE Approximate fractional statistical error in proton averageenergy
PAV Total proton energy flux (power density)--(B counter fluxrate times E).
APAV Approximate fractional statistical error in total energyflux.
282$
IAs an aid in data-correlation work, an arrow has been
placed at the bottom of these figures. A vertical line carries the po-la
sition of the arrow across all of the subfigures. The arrow shows the
point in time when the satellite crossed the L-shell that goes through
College, Alaska. Also indicated is a distance in kilometers--east or
west of College--where this transit took place. The 3-hour magnetic K
index from the College magnetic observatory of the U.S. C(Anstal and Geo-
detic Survey is also indicated.
All satellite paths for which data are used in this re-
port were north-to-south transits. We have indicated on the figures that
the left-hand side is north and the right-hand side is south as a reminder
of this fact.
4.3.2 Proton-Precipitation Data
Figures 4.D.1 through 4.D.14 present proton-precipitation
data as a function of orbital position. There are five subfigures in
these figure sets. The top subfigure provides as ordinate the logarithm
of the proton flux in the B counters. The B counters provide flux mea-
surements for protons with energies greater than 10 keV. The units are
protons per square centimeter per second per steradian. Regions in which
there are no data plotted are regions where the counting rate dropped to
a value to be expected from radioactive calibration sources alone. At
Lockheed's suggestion we have imposed a threshole belou; which no data
points were plotted.
The second subfigure from the top provides as ordinate
the logarithm of the average proton energy versus orbital position. The
ordinate is labeled E. The average energy is computed in kilovolts. The
average proton energy is estimated using the assumption that the proton-
energy spectrum is exponential in energy. Under this assumption the
29
relative counting rates in counters Band D can be used to estimate theI
I average energy. Note that for an exponential energy distribution the
average energy and the e-folding energy are the same.
The B counters and the D counters are independent of one
another. Each counter, measuring proton flux over a short time period,
provides statistically independent samples. Therefore, each value is
subject to statistical fluctuations; when the proton mean energy becomesI
large, or the absolute counting rate becomes small, the determination of
the average proton energy, E, becomes statistically uncertain. Based onI
information provided to us by Lockheed, we have produced in the third
subfigure an estimate of the statistical accuracy of the E estimate. In
fact, the way data are handled on board the satellite does not allow for
good statistical estimation of data accuracy; it is believed though that
A• provides an approximate fractional, statistical error in our estimate
of proton average energy. This is a linear scale of fractional error
from 0 to 1. Certainly wh~en our estimate of this error is close to one,
the estimate of • is quite uncertain. When the estimated error is nearI
0, then our estimate of average proton energy is liable to be .•tatisti-
cally accurate; howe•ver, the reader must remember that the derivation of
E assumes an exponential form for the proton spectrum. We do not know
S~whether this analytical form is a good fit to the real precipitation
!a spectrum.
! The fourth subfigure from the top presents the logarithm
! of the total proton energy flux as ordinate. The ordinate is labeled
!t PAV. This quantity, PAV, is the flux in Counter B times the average pro-
S~ton energy; hopefully, PAV is proportional to the incident power density
S~of protons impinging on the top side of the ionosphere.
The estimatc of PAV includes the statistically uncertain
estimate of E. We have therefore provided in the bottom subfigure an
S~30 iI
1-
Iestimate of the fractional error in the total proton energy flux. This
fractional, statistic error varies linearly from 0 to 1. The ordinate
is labeled APAV. As with our estimate of the fractional error in E,
when our estimate of the fractional error in PAV is large, our estimate
of PAV is clearly very poor. When the fractional error, APAV, is near
0. the accuracy of the statistical estimation of PAV will be quite good;
however, th, reader must remain cognizant of the fact that the quantity
PAV includes an assumption of an exponential proton energy spectrum.
As indicated above, we do not know how valid this assumption is. Table
4.2 summarizes the quantities in Figures 4.D.1 through 4.D.14.
As with the electron plots, the location of the satellite
transit of the College, Alask- L-shell is indicated, as is the College
three-hour magnetic index. Satellite transit is from north to south.
This is also indicated on the figure.
4.3.3 Histograms of Electron-Precipitation Data
In order to seek a definitive correlation between radar
aurora and electron precipitation a series of histograms has been prepared.
Data have been organized so that two histograms appear c each subfigure.
Each histogram presents the number of measurements of a particular quan-
tity from regions of the satellite pass where no radar echoes were ob-
served (i.e., regions where they could have been observed it the environment
had given rise to scattering). On tne same subfigure another histogram
presents the same quantity for regions where radar echoes were observed.
The histogram sets are given in Figures 4.E.1 through 4.E.14.
The top subfigure gives two histograms of the electron
|nuber flux--TSUM--according to presence or absence of radar aurora.
Note that both the abscissa and the histogram interval widths are loga-
Srithmic. For gqeater ease of display the ordinate is also logarithmic--
with a place piiovided for zero cases. Satellite data are read once per
31 *
- i: N
second. Therefore the number of cases plotted corresponds to the number
of seconds of data that were obtained, subject to other constraints. The
particle-flux threshold value of 5 X 106 was applied in order to eliminate
data dominated by the radioactive calibration source.
The middle subfigure gives two histograms of the
electron-energy flux (ELE) according to presence or absence of radar
aurora. Note that the abscissa and ordinate sre Jogarithmic.
The bottom subfigure gives two histograms of average
electron energy (EMN) according to presence or absence of radar aurora.
Only the ordinate is Logarithmic in this plot.
4.4 Comments on Specific Satellite Passes
In this subsection we discuss features of several of the satel-
lite passes that may help the reader to better understand the limitations
of this experiment. Some of these comments relate to work that is pre-
sented later, in Section 6 of this report.
4.4.1. Satellite Pass A388
If electron precipitation in the area of Bar 1 (see
Figure 4.A.5) produces radar clutter, then it probably would not be seen
by the radar since the usual clutter altitude of 110 km is just beyond
the radar horizon. Furth-rmore, the magnetic aspect is 30Y further dis-
criminating against observing clutter. The area of weak clutter shown is
bounded in the north by the southern boundary of detectable electron pre-
cipitation (threshold taken to be 5 X 106 el/cm2 -s-ster). There is no
evidence of proton precipitation in the area. An undetected electron
6precipitation flux in this area with a flux just below 5 X 10 could pro-
4 3duce 6 X 10 el/cm . This density might be enough to allow auroral radar
echoes to be obtained should this ionization become structured into
irregularities.
32
4.4.2 Satellite Pass A431
The maps of Pass A431 (Figure 4.A.11) show that the ob-
served radar clutter does not coincide with regions of maximum energy
deposition. The plots of the B-counter (proton) and the TSUM (electron)
count show that, in regions where clutter is observed, there is precipi-
tation of both protons and electrons. The E-re-ion %as not illuminated
by the sun. A monoenergetic electron beam of 3 or 4 keV, the average
energy indicated by EMN, would not penetrate to the appropriate altitude.
We presume that the actual spectrum is spread in energy. The proton
average energy exceeded 10 keV in this same area of the sky; this average
energy can produce ionization in the relevant regions of the ionosphere.
The electron flux in the main clutter region varies be-
tween 10 and 3 X 10 el/cm -s-ster. The proton flux is approximately
3 x 10 . The data of Figures 6.2 and 6.5 (see Section 6) suggest produc-
4 5tion of steady-state E-region densities of from 8 x 10 to 1.5 X 10 due
4to electron precipitation, and - 8 X 10 due to proton ionization. The
steady-state electron density due to both electrons and protons would5
therefore be likely to exceed 10 This background ionization is probably
sufficient to produce observable auroral radar clutter.
We note that the electron-precipitation flux is very
much greater, by a factor of 10 to 30, in regions where no clutter isA
obtained, though time simultaneity of measurement is not very good. Most
observers believe that the electron-density irregularities that cause
radar scattering from E-region ionization are caused by electric fields.
This single satellite pass may suggest that regions of higher precipita-
tion levels produce higher conductivities that cause "short-circuiting"
of the ionospheric electric fields.
The region of intense electron precipitation between 640
and 650 north latitude lies in an area where the magnetic aspect angle
33
from the radar exceeds 3". Therefore, the radar cannot well monitor the
clutter environment in that region of the sky.
4.4.3 Satellite Passes A2979 and A2995
These two passes occurred during a high-energy proton
aurora. The proton flux into the B counters was not extraordinarily
large, being in the neighborhood of 10 protons/cm -s-ster. But the
energies of the protons were so large that the determination of average
energy--assuming an exponential spectrum--was jeopardized by statistical
errors. That is, the B- and D-counter counting rates were nearly equal;
this means that most protons had energies above the 38-keV threshold of
the D counters. The ratio of the q to D counting rate would be nearly
1.0. The logarithm of this ratio is divided into a constant to determine
E. Any statistical fluctuation in counting rate of either counter could
bring the ratio to 1.0, thereby driving the logarithm to zero and causing
us to compute an infinite average energy. The proton energies are so
high that an "average energy" determination from the counting ratio of
B to D is statistically meaningless.
Electron precipitation was taking place nearly everywhere
on Pass A2979. On Pass A2995 the electron precipitation was also wide-
spread. The average electron energy was high on Pass A2979, being about
6 keV. On Pass A2995 the electron energy was somewhat less.
Radar aurora wus intense and very widespread during
Pass A2979. The clutter was weaker and contained spatial gaps on Pass
A2995. We could characterize both events as occurring during intense
aurora, though A2979 was stronger. In spite of the apparent tendency of
regions of peak precipitation intensity and radar aurora to anti-correlateiN(not overlap) during satellite passes when auroral activity is weaker,
there seems to be good overlap in the locations of the two phenomena for
these two passes.
34
Li SATELLITE 7 MIUE
rG
FIGURE 4.A.1 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. AND
¥- ~RADAR AURORAL DISTRIBUTIONS DURING PASS A518
35%
16 t
I&20
.1:0
WA4
A 4
RR~
FIGURE 4.A.3 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A624
37Iko
- ~ ~ ~ 0 :00~----
A 1A
A157
~2R.
cr KII
AZO~uTN-01STAME PROJECTON
IFIGURE 4.A-4 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT. -
PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT, AND 3RADAR AURORAL DISTRIBUTIONS DURING PASS A157
38
70.o
I AZAJTM-~StN~ POACT
I ~fi0:14:00FIGURE 4.A.5 MAP OF ALASKA SHOWING ROUDPOETINOtAELIEOBT
RADAR AURORAL DISTRIBUTION D~IN PS A
PRCPTTN6LCTO NRYFU6LOGTEOBT N
At" WEA
13 39Z
10160
74'a Is. lw S
TrIA~inJh-~SU~ M~CTM
10:51:00,
FIGUE 4..6 AP O ALSKA HOWIG GOUNDPROECTIN O SATLLT ORBIT
RADA AUORALDISRIBUION DURNG ASSA232
139 MI40 4ao
l~370 --rr IfAUOAI~RPS
0:38:0Io
10390IA
NO AURORA ON 139OR 398 MHz AT SATELLITE TIME
aZwmj-os. ~TAIcZ PROACTOVrnma .Dit
FIGURE 4.A-7 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PREC.PITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A676
41
AtI
|I
£ZE 6 MAY 1969--
11040 &ISM• 398 MHz
0 ono Nma
FIGURE 4.A.8 MAP OF AI.ASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT, ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A751
42
aA
_~1270 10-:0.!F
~~0A
RADA AURRAL ISTRBUTINS DRINGPAS 1396Este DoteýBlaire Lj'keA36
43!,"',13 ~
- ~ ~-----ELECTRON-
I
•*1
,10O-51:OO0I a II
10:53:00 •
S\30 APRIL 1969
1 e
.Z.-
FIGURE 4.A,10 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON 4.NERGY FLUX ALONG THE ORBIT, ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A657
44
( __
'.1.awl
r4*w
Al
\1 OCTOBER 1969
PRECPITAINGELECRONENERY FLX AONG HE RBICT.AN
RADAR AURORAL DISTRIBUTIONS DURING PASS A2995
47
Go- 4 -5
VIISUAL
A
NN
I 30 SEPTEMBER 1969
1-30 -139 Mhlz (398 ~ is ffmuaha)
404?
FIGURE 4.A.14 MAP OF ALASKA SHOWING GROUND PROJECTION OF SATELLITE ORBIT.PRECIPITATING ELECTRON ENERGY FLUX ALONG THE ORBIT. ANDRADAR AURORAL DISTRIBUTIONS DURING PASS A2979
48
; .~
-110 -100 -JO
-144-19
FIGURE 4.8.0 CALIBRATION OF RADAR A-SCOPE TRACES FOR USE WITH
FIGURES 4.B.1 THROUGH 4.B.14
3949
a.
00
In
to U)
F4 w 00 0j
00
4: 0
co 0C) LL
>D U, >
4- u
0 Im tN 0 0n 0
o c 0r< 0
> C.) w.
LLa
2
c~,50
oP
L i !N
C-i~
4 ~ ~ 00
> CUoU
0T <
0 0
z
0 0 0
>w
-j CP
I- s 0
4U cn
0 0
In On
z4
- 0
z B mI4F M
UJK w
I::C.,
51
L 0~
-
- 0i
0*--
N<
00 LU- U
zz z c
o2 "
0. cm 0 r 0 U
Q - r
I. C w
LU LUC-i -i U,LLJ L0
cc)
4 0 I- KV 00
UJ 52
13 M(z 139
L111'LI< AI
SRANGE: -- km 144-18
•. ~12 APRIL 1969 -
"•, • FIGURE 4.B.5 RADAR A-SCOPE RECORDS FOR SATELLITE PAS;S A388
454
AZIMUTHi~rs2rr..o
_ _ _ _____ _ _
SAZIMUTH 3 ELEVATION 3' AZIMUTH 30 ELEVATION 3' AZIMUTH 3" ELEVATION 3'/4'
38139 MHz 139Mz
~Ifl
1050 00-105"30 1050.00-10500 10500-1051:0
Si I I t I I II I I I I
AZIMUTH 3' ELEVATION 4' AZIMUTH 3' ELEVATION 4' AZIMUTH 3 ELEVATION 4l
ELEVATION 4 AZIMUTH 3 ELEVATION 44 d
S139 M2Az 13913
A 0 00 50 050 5 10 50 05010 W
FZIUTH3'ELVTO 4 RAARI A-SCOPE ECORDS TFOR S^ATELLITE PAS ELEVTIO 4'
I0H~139 139MH1 55
IIA I.-
.8.
C-l,
ww
o So0 C0
Ca
LU)
LU cc0
z cc
00 ~C.
L <,
-~ C3I- C<
0 4U-j ccC-, I
Z *L
4??mo
o 70
3anil-dM 90
56
--------- .....
"AZIMUTH "V ELEVATION 2' AZIMUTH "•, ELEVATION 2' AZIMUTH 14' ELEVATION 'V
398 M~~z 398 MHz 38~
1100-5G-1 101:20 11301020 1020- 1102-'nSIII I I I I III
AZIMUTH 12 ELEVATION - AZIMUTH 12' ELEVATION 2' AZIMUTH "- ELEVATION 2'
0~4
S1103:00-110:30 03"30-1104 001102"30-1103:00
I I I I I I I I I
AZIMUTH - ELEVATION 2 AZIMUTH 'V ELEVATION 2' AZIMUTH ELEVATION 2 7'
1104:50111135.1104 1105 75 1105 55
tI I I I I i "I I I II
0 500 1000 1500 0 500 1000 1500 0 500 1000 1500
RANGE - -km R o14d-96 MAY 1969
FIGURE 4.B.8 RADAR A-SCOPE RECORDS FOR SATELLITE PASS A751
57
5 -
j; j0 C
o o c04 C) C3
Cd -,
C--J
on 8 tov
2 2ow 0
* w~ w
z 0
0 0 oa
> IU2i E
w 4 *Uz' <
0i0CIO
DD 0
0 - 0
-~~~~ NJOl Ii I ~
59l
I_<
....... ... ... ... ......
-M127OD127:0-E12730-11228:0
E1228:00-1228:'o E1228.30-1229-01
J I II III IF0 500 1000 1500 0 500 1000 1500
RANGE - km 146-23I29 SEPTEMBER 1969
FIGURE 4.13.12 RADAR A-SCOPE RECORDS FOR SATELLITE PASS A2964
61
AA
IW 15:0-15 1154'30-1155'00
="1155-00-1155-: 1 0ý
IN -i 1
I OCTOBER I 96S T
000000m 1500:O-1543 150301150'0 0
RAGE -- k-14-
C% FIGURE 4.1.13 RADAR A-SCOPFz 13CORS FOR SATE-LLiTE PASS A2995
in
-
U-U
1211 112
1212 00-12123 1212 30-1213.00
0 Soo 1000 1500 0 500 1000 1500
RANGE - kni 147-8
30 SEPTEMBER 1969
FIGURE 4.13.14 RADAR A-SCOPE RECORDS FOR SATELLITE PASS A2979
63
,. ~ ~ w rW -- - -
I T 1010
T-P
SUM
106 . _ _ _ _
e I-1 0 .. ."
0 I
1010
ELE
106 A518
NORTH 10.14:00 10:15:00 10'16:00 10:17:00 SOUTHI 260 km -- E
FIGURE 4.C.1 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN)o AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A518
64
:1 ___
1010
1116
10
EMN
0
1010
ELEA'
BEYOND106 HORIZON __ _ _ A641
NORTH 11:05:00 11:06:00 11:07:00 11:08.03 SOUTH
240 km WK -0
FIGURE 4.C.2 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN), AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A641
65
Iý
1010
T-P
106o
10 I
EMN
ELE A
106 A624
NrJORTH 10:13 00 1r.14:fl0 10:15.C" 10:1C 00 SOUTH
3r.0 kr .- EK-I
F;GURE 4.C.3 PRECIPITATING •LECTRON NUMBER FLUX (TSUM), AVERAGE
ENERGY (EMN). AND ENERGY FLUX (ELF) versus UNIVERSALTIME FOR SATELLITE PASS A624
66
1010
T-Psum
106 1
ENIN10
1010 . -
ELE
106 LA157
NORTH 10:20:00 10:21:00 10:22.00 10:23:00 SOUTH
S0 km
Ki
FIGURE 4.C.4 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOP SATELLTE PASS A157
67 j6
NI
'. 1010
T-PISUM
106
10
EMN
N ~0 -
SHORIZON
A333• 106
NORTH 10:13:00 10:14:00 10:15:00 10:16:00 SOUTH210 km- E
FIGURE 4.C.5 PRECIPITATING ELECTRON NUMBER FLUX (TSUM). AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) vet UNIVERSALTIME FOR SATELLITE PASS A388
68
-' *1
1010
EMN
SUM s 1 •3
106 A232
NORTH 10:52:00 10:53:00 10:54.00 10:,55-.00 SOUTH
280 km -- W
1K - 2
FIGURE 4.C.6 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSAL
01 1
_TIME FOR SATELLITE PSA2
NOTH1052O~ 1:5:0 1:5.0 1:5:06SUT
1010
T-PSUM
106
EMN•
ELE
I ____ __A676
NCRTH 10.38'00 10.39'0,r 10:40.00 10:41:00 SOUTH
100 km -- EK 2
FIGURE 4.C.7 PRECIPITATING ,.' :CTRON NUMBER FLUX (TSUM). AVE, -%GEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A676
70
1010
T-PSUM
106
lO
1010
ELE 17j
106 _____ A751
I I !
NORTH 11:0t0 11:04:(0 11.05:00 11:06.00 SOUTH
150 km- WK - 3
FIGURE 4.C.8 PRECIPITATING ELECTRON NUMBER FLUX (TSUMI. AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A751
1-<
71
1010
i,-P I
SUM
106
'o r
EMN
0 2
,o o I
ELE -- Ai .\
.06 2I !I I•NORTH 10.27:00 10:28.00 10"29.00 10:30:00 SOUTH
0 kmK - 4
FIGURE 4.C.9 PRECIPITATING ELECTRON NUMBER FLUX (TSUM). AVERAGEENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A369
72
10 10
T-P h
1061 _ _
10
10 10
ELF
106 I10
NORTH; 51:00 10:52:00 10:53:00 10:54:00 SOUTH
ENERGY (EMN). AND ENERGY FLUX (ELE) versus UNIVERSAL
4 TIME FOR SATELLiTE PASS A657
734'
loin
T-P
SUM
10
EMN hV0 01
NORTH 11:08.00 11,09.00 11 "0.00 11 :11.00 SOUTH.,"
370 km - W
FIGURE 4.C.11 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVE.RAGE
ENERGY (EMN)o AND ENERGY FLUX (ELE) versus UNIVERSAL •
S~TIME FOR SATELLITE PASS A431
744
- - --
T°PJ - . j .- , •-
1010 1
T-P
SUM
106 __
10-
; ~FIGURE 4.C.12 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGE =ENERGY (EMN), AND MNERGY Fl UX (ELE) versus UNIVERSAL
STIME FOR SATELLITE PASS A2964 "t
775
ij
, P{
liii
T-PSUM
106
10
EMN 1
1010
ELE
1•06 ................ } JA2995
NORTH 11:54:00 11:55:00 11.56:06 11.57:00 SOUTH
530 km- Ei ' K' 7
FIGURE 4.C.13 PRECIPITATING ELECTRON NUMBER FLUX (TSUM), AVERAGEENERGY (EMN), AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A2995
76
I _ _
1010
T-P A .SSUM ....
-- um
106
"•110 .- V IEMN
1010 ~-
ELE
fi;YOND1"O N A2979S~106 .
NORTH 12 10:00 12:11:00 12:12.00 12:13:00 SOUTH
300 km - E
FIGURE 4.C.14 PRECIPITATING ELECTRON NUMBER FLUX C. (,UM), AVERAGE
ENERGY (EMN), AND ENERGY FLUX (ELE) versus UNIVERSALTIME FOR SATELLITE PASS A2979
-- .7
Slos1
105
103
E
101
PAV
106
t, '•APAV ]-
4 I-] A518
,II I
NORTH 10:14:00 10.15:00 10:16:00 10:17:00 SOUTH
260 km- E
K -0
FIGURE 4.D.1 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (Ei).
STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A518
78
i4
?Nf~s
HIGURE 4.D.2 PRECIPITATING PROTON NUMBER FLUX (B), AVERAGE ENERGY (E),STATISTICAL ERROR iN AVERAGE ENERGY (AE'), ENERGY FLUX (PAV),
AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A641
7
,i
103
B
l1f\
10-lop
',103 1.•
101
• • :A6240
K - 1
I PAVI I I8I
NORTH 10:13:00 10:14:00___ 10:15.00 10:16:00__ SOT
iIPA km--
FIGURE 4.D.3 PRECIPITATING PROTON NUMBER FLUX '.B). AVERAGE ENERGY (E'),i
STATISTICAL ERROR IN AVERAGE ENERGY (tiE), ENERGY FLUX (PAV),
AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A624 ;
80 :
04
10" -
-El
g.I010..
. . - I
FIGURE 4.0.4 PRECIPITATING PROTON NUMBER FLUX ((B). AVERAGE ENERGY (E).
STATISTICAL ERROR IN AVERAGE ENERGY ME•). ENERGY FLUX (PAV),AND STATISTICAL ERROR IN ENERGY FLUX (A•PAV} vemus UNIVERSAL :TIME FOR SATELLITE PASS A157
1081
108
Eu
.. 1
lo,
AE1
InPAV
HORIZON . II I __ _
NORTH 10:13:00 10:14:00 10,.15:00 10:16:00 SOUTH210 km- E
K- 1
FIGURE 4.D.5 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E).STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STAIrSTICAL ERROR IN ENERGY FLUX (APAV) veris UNIVERSALTIME FOR SATELLITE PASS A388
82
i:Qo -0a-"aa
1031
- -A
E
10" "
lop
APAV
I o II•
NORTH 10:52:00 10:53:00 10:54:00 10:55:0 SOUTHa
280 km -W
K - 1
833
S~FIGURE 4.0.6 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E').
• -•-7SThTISTICAL ERROR IN AVERAGE ENERGY (AE'). ENERGY FLUX (PAV).
S~AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALS~TIME FOR SATELLITE PASS A232
83 w
10:3 i
10.
E
101
idPAV
1 9I
APAV
L ,A676
NORTH 10.38:00 10:39:00 10:40:00 1C:41.CO SOUTH
100 km - EK-?
FIGURE 4.D.7 PRECIPITATING PROTON NUMBER FLUX (B,. AVERAGE ENERGY (E).STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV). versus UNIVERSALTIME FOR SATELLITE PASS A676
84
-•-
lopl
- ----- IA °10•
i t If
A75
EN
1050 k - 1
_ _ _ _ _ K -13ARPAV _
TIM FOA7TLLT5PS A5
! !j JINORTH 11 03:00 11:06:00 11:015:00 11:06:00 SOUT.'
150 km -- W
• ~FIGURE 4.0.8 PRECIPITATING PROTON NUMBER FLUX (B). AVEP•AGE ENERGY (E),
I STATISTICAL ERROR IN AVERAGE ENERGY (AE'), ENIERGY FLUX (PAV).
AND STATISTICAL ERROR IN ENERGY FLUX (&PAV) versus UNIVERSALTIME FOR SATELLITE PASS A751
iVa-
loý
103
lop
105 . ..
PAV
106
NORTH 10:.27:00 10:28:00 10:29:00 10:30:00 SOUTH0 km
FIGURE 4.D.9 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E).
STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV) verss UNIVERSAL
1086TIM FR ATELIE ASSA39
____I
10:30:0 SOUT
IdIW
10
H ON
1 106
f" H -1:0 10520010530 10:40 OT
APAV~ -' 5
FIGURE 4.D.10 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY ( E).STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAy).AND STATISTICAL ERROR IN ENERGY FLUX (.APAV) versus UNIVERSALTIME FOR SATELLITE PASS A.657
87
lop
los _ _ _ _ _ _ _ _
01
AN TTSTCLERO NEERYFU VAAV veu UNIVRA RoP______________TIME____ FOR SAAGNETTC ASPEC A431
NORH 1:8:0 1:9:Q 1:1:0 1:1:0 SUT
FIGURE 4.0.12 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY OF).
STATISTICAL ERROR IN AVERAGE ENERGY (CE), ENERGY FLUX (PAV),
AND STATISTICAL ERROR• IN ENERGY FLUX (APAV) vresus UNIVER•SALTIME FOR SATELLITE PASS A2964
89 •
;4
108 ,'..,
'A. IC:^ • ' '
- . I%
1013..
AE
106 "
PAV
APAV "
01 ~ A2995
NORTH 11:54:00 11:55:00 11:56:00 11:57:00 SOUTH
530 km - E
FIGURE 4.0.13 PRECIPITATING PROTON NLAMBER FLUX (B). AVERAGE ENERGY (E),STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV).AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A2995
90
* o's .. . . l .. -
103
IL
I 1lol HOIZN
10 6
APAV~ ~HORIZ ON i AM 9'7o I I
NORTH 12:10:00 12:11:00 12:12:00 12:13:00 SOUTH
-W km- EKt-7
FIGURE 4.D.14 PRECIPITATING PROTON NUMBER FLUX (B). AVERAGE ENERGY (E),STATISTICAL ERROR IN AVERAGE ENERGY (AE). ENERGY FLUX (PAV),AND STATISTICAL ERROR IN ENERGY FLUX (APAV) versus UNIVERSALTIME FOR SATELLITE PASS A2979
91
7; 1 _
. 100
< ->
. 10
U)
BELOW 3-1106 le 108 091010THRESHOLD TSUM -- i-2 S71 ster'.
LU1a I I
100U)
z -
wa - - ,;N.Z 10U)
0II
BELOW 31100 10e 108 109 1010
+THRESHOLD ELE - keV/cm2 s ster
.~J1 0 0 IRADAR> l "-AURORA PRESENT
I.- NO RADAR; 10 AURORA PRESENT
C" A518
<*~*'K - 0
1 -~SimnultareitV 7 minutes
0 5 10
EMN - keV
FIGURE 4.E.1 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITAT!NGEL.ECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A518
92
. 100
> g
z S10"J •~.j .
BELOW 3x10 6 10e 108 109 1010
THRESHOLD TSUM - cm- 2 s-1 ster-1
j 100
"C
o I I.BELOW 3x108 lO 108 109 O101
THRESHOLD ELE -- keVlcm2 s ster
,100 I-RADAR
• -- _AURORA PRESENTw -iI- -- -- NO RADAR
-_ 10 • AURORA PRESENT"U)
0A541
THEHL 5 L 10/c S2taet -1 minute
EMN - keV
FIGURE 4.E.2 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATINGELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A64I
- - _ ~ -4 10
-j 100
cc
z. 10
In
0
BELOW 3x106 107 108 109 1010
THRESHOLD TSUM - cn- 2 s 1 Ste.-i
100
wI-.
z
0
BELOW 3x10 6 107 108 1010
THRESHOLD ELE - keVlcr 2 s $teR
j100 I_j 10 __-- 8-"'7 RADAR
:> F-1 AURORA PRESENT
t- •• IL•-•r,•'• NO RADAR
10 "AURORA PRESENTIn
K -•" ,•,• . •Simultaneity -- Good
0 10
EMN - keV
FIGURE 4.E.3 HISTOGRAMS OF LOGARITHMIC OCCURRENCE veasus PRECIPITATINGELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A624
94
-; --
AI
Z 100
10
SBELOW 3.106 1oe 108 log 1010THRESHOLD TU-& 1 s&
:>cc
:100
z< 10
U)
w
BELOW 3.106 107 108 log 1010
SELE - keVkcm 2 s ster
_j 100 •_.•< -- 0 RADAR
> -- AURORA PRESENT-- -NO RADAR 75
- 10 AURORA PRESENTU)
~ ,~-,A157
U) K ='2A]4 ~., .' Simultaneity - Good
010 5 10
EMN - keY
FIGURE 4.E.4 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATING J;
ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A157
95
0 0 ~~~w -A;
-'S
-J 1)00-
z
,,g,
'- -- -
10..,
0 . ..... "" k ........*'•':
BELOW 3.10 6 1(7 1o8 19 10 10
THRESHOLD TEUM - c-cm2 s-1 stef1
"J100
AURORAPRESEN
I- -- NO RADARz -- • AURORA PRESENT
.ir
10 ~~
4 1 •;•• -••, ••, - Simultaneity -- Good
00 5 10
EMN - keY
FIGURE 4.E.5 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATING
ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A388
96
> F- UOA RSN
- -u
100
ILl
0BELOW 3x10 6 107 108 109 1010
THRESHOLD TSUM - cm-2 s 1 ster"1
-j 1004>
z10
CL)
BELOW 3x 06 107 108 log i010THRESHOLD -ELE -- keV/cm2 $ ster
S100 R A D A> - F AUROR•A PRESENT
W-•10 • • AURORA PRESENI[
CL)THRE-HOL ELE-NOy/RADARte
-K
rn Simultaneity - Good
S~0 00 5 10
EMN - keV
FIGURE 4.E.6 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATINGELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A232 N
97
1001
> -
Z 0
0-
BELOW 3.10 6 10 i8 109 101THRESHOLD TU m 2 slse-
I100 RA R
>JARR RSN
NORAA
Siutaet -Go
FIUE10UOA RSNOCURNE essPECPTTN
ELCRNNME LU TU) NRY LX(L) NAVRG.NRY(M)FO AELT ASA7
U98
,j 100 -
10lc--
w~ 10
BELOW 30106 10e 108 109 1010
THRESHOLD TSUM -cm-2 7-1 ster-1
100A
• z
4 0
0 °
BELOW 3x`10 107 108 10 101
THRESHOLD ELS - keVlcm2 s ster
100
> F1AURORA PRESENTcc
4/
-- - |•'P0 NO RADAR. Z10 • • AURORA PRESENT
to
K -
~ 0
0 5 10A
EMN - keY
FIGURE 4.E.8 HISTrOG3RAMS OF LOGARITAMIC OCCURRENCE versus PRECIPITATING
ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PAS A751
99
wýj
.j100
>
cc
z -I I q•! I ,
BELOW 3x106 107 108 log 1010
THRESHOLD TSUM -- cm-2 s-1 ster-1
I-2
.100
- -o<4 I l I
>ui
-- .-n.*___BELOW 3x10 6 107 108 109 101
THRESHOLD " 4 : cO)tC ~ j RSNll
0IBELOW 3x10 6 10 181911ELE - keV/cm s ster
J100S1I-RADAR
S1 AURORA PRESENT
S1o • •:• • NO RADAR10 , . AURORA PRESENT
"CL Y •y.. A369
z Simulitaneity-i minute
010 5a-L 10EMN - keV
FIGURE 4.E.9 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATING
ELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A369
100
100
-I100
w
I-z10
w
0.
BELOW 3.106 107 108 109 1010THRESHOLD TSUM - cm"2 s- 1 ster-
-J100
S10
BELOW 3-10 107 108 109 100TELE -- keVcm2 s ster
z <'1°°" _- [ RADAR: > __AURORA PRESENT 2
U)
I- "- -- NO RADAR
10 A P
ELE K -k592
300
FIGURE 4.E.10 HISTOGRAMS OF LOGARITHMIC OCCURRENCE veADAs PRECIP;TATING
ELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), AND
AVERAGE ENERGY (EMN) FOR SATELLITE PASS A657
Szoz
101i
00 ~Samuta~ety -Goa
• - , I i'• " •;.. .+ +"+ '" • " ',;+ '+' •:'+ +-".• •+ .+m .- • +:.-•: +-.,,•..+ ... ., .. + .. +++.+.+ • +.10+.
j 1004c
7- 1Ion
0
BELOW 3x10 10 i8 109 100THRESHOLD TSUM -cmn
2 S71 ster-1
100
zZ 10-
0-
BELOW 3-io 1 _C I07181911THRESHOLD ELE keVlcm2 s ster
100 RADAR
NO RADARz- 10 AURORA PRESENT
C--"d ~A431
In~ = K5'14M1 Simultaneity -3 minutes
00 510
EMN -keV
FIGURE 4.E.11 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATINGELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A431
102
S100>
z '
0
BELOW 3.106 10 log10 1010THRESHOLD TSUM -fj- S7 ster-I
4100
1 0
BELOW 3-106 10 109 10100THRESHOLD ELE -- keVlcm2 s ster
.j100 RADAR<~ AURORA PRESENT
Lu~ NO RADAR10 AURORA PRESENT
A2964~' K - 6
~, ~Simultaneity -Fair
00 5 10
EMN - keV
FIGURE 4.E.12 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIP!TATINGELECTRON NUMBER FLUX (iSUM), ENERGY FLUX (ELE). ANDAVERAGE ENERGY (EMN;t FOR SATELLITE PASS A2964
103
j 100a>
::z ~ 100
LU
C
I-.
IBELOW 3.c-§ 10e 108 109 1010
THRESHOLD TSUM - cm -2 s -1 ster -1
.j100 -
-10
u,
-F
BELOW 3-106 107 108 lop 1010
THRESHOLD ELE - kecm2 s' ft- r
.<100 _ -I- • RADAR
•> AURORA PRESENT
Sus
uJ I • •-- • NO RADAR-- 1zr-' AURORA PRESENT
S1io iutoet •• oCL
10toK=
0 5 10EMN - L kcV
FIGURE 4.E.13 HISTOGRAMS OF LOGARITHMIC OCCURRENCE versus PRECIPITATING
ELECTRON NUMBER FLUX (TSUM). ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A29
100
tr 11-10tr
0.-
•N -, R.
BELOW 3x106 107 0 109 100
THRESHOLD TSUM -- cm-2 s-1 ster-1w 100
0-0. - -- - I-
BELOW 3x10 6 107 108 10 1010
THRESHOLD ELE - cm 2 st
< -- -- RADAR> - AURORA PRESENT
z -- ~ NO RADAR
S•.,,••:Sinultaneity -- Gnod
0 5 10
ELECTRON NUMBER FLUX (TSUM), ENERGY FLUX (ELE), ANDAVERAGE ENERGY (EMN) FOR SATELLITE PASS A2979 _
105
....... ......
5. COMPARISONS OF RADAR AURORA
WITH MAGNETOMETER AND ALL-SKY-CAMERA DATA
In this section two other kinds of data correlations that may shed
light on the understanding of radar aurora are presented. The first
study is a comparison of magnetometer data with radar aurora. The second
comparison is between radar aurora and visual aurora as recorded with
all-sky cameras.
5.1 Radar-Aurora Magnetometer Comparison
On 24 March 1969 radar data were collected for approximately
5 hours for the purpose of seeking a correlation between radar echoes
and magnetometer response for a magnetometer placed below the region from
which radar echoes were being obtained. The radar antenna was directed
toward the ionosphere at an qltitude of 110 kilometers on the magnetic
field line that intersected a magnetometer operated by the University of
Alaska located at College, Alaska. We present here a composite figure of
radar data and magnetometer data; the composite is given in Figure 5.1.
The upper portion of the figure gives magnetometer response versus Green-
wich Meridian Time for the Z and H compinents of magnetic field, At the!t
bottom of the field are three amplitude-versus-time traces. The ampli-
tudes are given in dBm. These are radar signal amplitudes obtained witb
the 139-MHz radar. The amplitudes were measured using a 40-km range gate.
The rangc gates were centered on the magnetic field line through the
magnetometer (center trace) and at ranges less than and beyond the mag-
netometer field line--as indicated in the figure.
Visual inspection of the data presented in Figure 5.1 does not
show an obvious correlation between behavior of these two instruments
porECEWNS PACE BAK 107I
I____
(radar and magnetometer). The magnetometer samples ionospheric currents
over a fairly large north/south extent of the ionosphere. It is probable
that our three radar range gates are so close together that the magnetom-
eter averages over effects in all three gates. It must also be remembered
that magnetometers respond to current in the ionosphere; it is believed
that conditions for the structuring of ionization that leads to radar
scattering requires ionospheric electric fields, not electric currents.
Therefore lack of an obvious, detailed correlation may not be surprising.
5.2 Radar Aurora--All-Sky-Camera Comparison
The question of relationships between visual aurora and radar
aurora has not been unambiguously resolved. There is a tendency, we
* believe, for radar aurora to be located near to but not necessarily coin-
cident with visual aurora. This relationship has been found on occasion
during periods of fairly stable, quiescent aurora. The study of the re-
lation of radar aurora to visual aurora currently requires stable, well
defined visual arcs and stable radar aurora. This is because it takes
several minutes to accomplish radar scans to map radar aurora. Many
aurora will move very great distances during this period of time. It is
expected that with the addition of the phased-array radar to the Homer
facility, more dynamic aurora can be investigated to seek visual radar
relations.
Our experience has shown that radar/visual correlations are
only meaningful if the visual data are obtained by triangulation between
at least two camera sites. For Alaska this requires inordinately good
weather. Otherwise, cloud cover prevents successful acquisition of all-
sky data.
Figures 5.2 and 5.3 provide two examples of radar visual corre-
lations. These data were obtained with the 139-Mfz radar. The data for
108
Figure 5.2 were obtained at 12:19 GMT on 24 September 1969. The data for
Figure 5.3 were obtained at 12:28 GMT on 10 October 1969. In Figure 5.2,-J$
two visual arcs are indicated; these were located by triangulation between
all-sky photographs taken at Ft. Yukon and at College, Alaska. The re-
gions of radar return are indicated by the bars located generally to the
north of the northern visual arc. We note that the radar returns seem
to be adjacent to, but not superposed upon, the visual arc. We note that
in the eastern end of the northern visual arc, the visual arc seems to
bifurcate the radar echo. The data of Figure 5.3 generally show radar
returns that are also bifurcated by the visual arc. '7
It has been speculated2 that since radar aurora is thought to
be produced by high electric fields, the high conductivity produced by
ionization generated by the precipitating particles that cause visual
arcs will short circuit these electric fields. In the absence of the
electric field, plasma instabilities cannot grow to produce structured
ionization that is needed for radar scattering. These visual/radar com-
parisons seem consistent with this speculation.
109
109
6. STUDY OF SATELLITE DATA
It is now universally agreed that radar aurora is produced by scat-
tering of radar energy from field-aligned irregularities in E-region
ionization. There are several important questions that are continually
being addressed. Some of these are as follows:
(1) What produces the E-region ionization (during nighttime)?
(2) What mechanisms produce irregularities in the E-region
ionization?
(3) What is the detailed relationship between radar aurora
and other auroral phenomena?
In this section we study some of the physics and chemistry of pre-
cipitating particle interactions in the ionosphere as background to in-
terpreting the satellite data. We then concern ourselves with preliminary
interpretation of the data presented in Sections 4 and 5 as a start in
clarifying our understanding of the problems posed as questions above.
6.1 E-Region Ionization Produced by Precipitating Electrons
A question of concern in radar aurora is whether precipitating
particles produce the ionization that is necessary for radar clutter
echoes when the E-region is not sunlit. Daytime (sunlit) radar aurora
can be produced by structuring of normal daytime E-region electrons.
We have pursued this question in relation to the satellite precipitation
data in order to better understand the production of nighttime E-region
ionization required to produce auroral radar returns. The data provided
to rs by Lockheed only gave the average energy of precipitating electrons.
113
The detailed spectrum based on their five energy-selecting channeltron
counters was not made available to us. Figure 6.1 presents dataP giving
the altitude of maximum ionization rate due to precipitating electrons.
The data in this figure are for precipitation at a geomagnetic latitude
of 670. The pitch-angle distribution is assumed to be isotropic over the
downward hemisphere. The data of Figure 6.1 are for a monoenergetic
electron energy spectrum.
* The precipitating electron spectra are not monoenergetic. The
proton data to be presented later, as well as the rate of photon-production
data for an exponential electron spectrum,7 reveal deeper penetration
than indicated in Figure 6.1 if the altitude of the effect is referenced
to average particle energy. That is to say, had we used an exponential
spectrum for the electrons and plotted ionization rate versus average
electron energy, then the curves in Figure 6.1 would be moved to a lov.er
altitude.
Figure 6.1 indicates by cross-hatching the band of altitudes
over which radar aurora is most frequently observed. For monoenergetic
electrons with isotropic pitch-angle distribution, energies must exceed
5 keV before ionization is produced in the altitude region of prime
concern to radar clutter. Because electron spectra probably have a spread
in energy when the EMN from the channeltron spectrometer exceeds 2 kV
or so, we presume that ionization will be produced by the electrons in
the altitude region of the most frequent and most intense radar clutter.
Since the satellite data provide an absolute flux, it is pos-
sible to estimate the E-region electron density produced by the precipi-
tating particles. Figure 6.2 provides some results of these computations.
Altitude in kilometers is plotted versus ionospheric electron density for
various precipitating electron fluxes. These electron densities are
computed at each altitude using a monoenergetic electron beam that pro-
duces its maximum ionization rate at that altitude. Thus the computation
114
J-
at 100 km altitude assumes a monoenergetic electron flux of 17 keV %see
Figure 6.1). The calculation for an altitude of 130 km., for example,
assumes an electron energy of 2.5 keV.
The steady-state electron density in these calculations was
computed assuming that unperturbed atmospheric chemistry is pertinent.
The electron clean-up mechanism that dominates the chemistry is disas-
sociative recombination. This reaction varies with altitude, according
to data summarized elsewhere. 8 The reaction is also temperature-sensitive
so that if the auroral precipitation or the auroral electrojet heat the
ionosphere above normal temperatures, higher electron densities than we
compute may be produced.
The radar clutter data do not directly reveal the electron
density in the scattering regions. Under very special circumstances one
can deduce approximate values for this quantity (see Ref. 2). When this
is done, values of nighttime E-zegion electron densities between 10 and6 3
106 el/cm are found. To produce these densities, accorcing to Figure7 2
6.2, fluxes equal to or greater than 10 el/cm -s-ster are required. The
electron counter threshold, below which Lockheed advised us not to use6
the data, is 5 X 10 . Figure 6.2 indicates, therefore, that when electron
precipitation is measured by the satellite above its detection threshold,
the precipitation is nearly capable of producing the electron densities
that seem to be needed.
Another important question is that of persistence of the ioni-
zation produced by precipitation. If the precipitation suddenly turns
on, how long does it take until the ionospheric electron density reaches
the steady-state electron density as determined from Figure 6.2? Figure
6.3 presents altitude versus chemical time-constant for the electron-
precipitation fluxes described by Figure 6.2. In the altitude region of
concern (105 to 115 kilometers), Figure 6.3 indicates that chemistry
115
lifetimes are shorter than 20 to 30 seconds for the satellite electron
precipitation threshold of 5 X 106. We conclude from this figure and
Figure 6.2 that the electron densities produced by electron precipitation
as measured by the Lockheed satellite are great enough and can be pro-
duc~d in a sufficiently short time to explain the presence of electrons
at E-region altitudes during the nighttime sten auroral clutter is ob-
served. We study Later whether there is spatial overlap of ineasureable
electron precipitation and radar aurora.
6.2 E-Region Ionization Produced by Precipitating Proton
As with electron precipitation we ask the question, do precipi-
tating protons produce enough E-region ionization--in the absence of
sunlight--to explain radar auroral clutter when clutter and proton pre-
cipitation spatially overlap? To investigate this we have plotted data
obtained from Chamberlain. 9 The data are for an exponential (in energy)
proton spectrum, and an isotropiv pitch-angle distribution in the downward-
going hemisphere. The results are presented in Figure 6.4. Also indi-
cated is the altitude at which the ionization production rate is 10 percent
o0 the peak value. The upper-altitude 10-percent region is so high it
was not plotted.
To aid in ,mderstanding penetration of a continuous spectrum
compared with the monoencrgetic spectrum that was used for tthe electron
£ precipitation, the stopping altitude for monoenergetic protons with zero
pitch angle was also plotted in Figure 6.4. The isotropic pitch angle
distribution would tend to raise the altitude of peak ionization with
respect tc, this lntter, zero-pitch-angle curve. The high-energy tail of
an exponential spectrum moves the altitude of peak ionization rate do*rn-
ward into the atmosphere, if one plots peak ionization rate versus average
proton energy. We note that the proton average energy must exceed about
116
10 keV in order to produce ionization in the altitude band of most fre-
quent and intense auroral radar clutter.
As we interpret the data provided in Ref. 9 we can only compute
the average electron density produced by a proton exponential energy
spectrum with average energy at 32 keV. Using the reaction rates of Ref.
8 we have plotted the electron density versus proton flux in Figure 6.5.
We assume that a background ionization density of 105 el/cm3 is needed
to explain auroral clutter; Figure 6.5 indicates that the proton flux
vould have to exceed 5 X 10 in order to produce the needed ionization in
the absence of solar (daytime) E-region ionization.
It is important to note that precipitating protons can undergo
a process that is not available to the electrons. A precipitating proton
can pick up an electron from an atmospheric atom and thereby become a
neutral particle. Once the proton is neutralized it is no longer con-
strained to follow the earth's magnetic field lines. On deeper penetra-
tion of the atmosphere, the neutral proton may again become ionized.
By this process the protons can migrate across magnetic field lines.
Therefore, proton fluxes that are measured by the satellite at altitudes
of about 400 km may be spatially spread by charge exchange by the time
the proton reaches altitudes of 100 to 120 1m. As a result, a localized
peak in proton flux measured at an altitude of 400 km may become spatially
spread by 100 km or more by the time the protons reach E-region altitudes.
Estimates of E-region ionization that may be produced by proton precipi-
tation as measured by the satellite must be spatially averaged along the
satellite path in some manner. We have not pursued this question in any
iaore than qualitative depth at this time.
117
4
41
6.3 Correlation of Radar Echoes with Satellite-Borne Particle-
Precipitation Measurements
In this subsection we compare some results of the simultaneous
measurement of radar aurora and particle precipitation. In particular,
we compare locations of peak energy flux with locations of radar aurora.
In some sense, this kind of measurement may be redundant with radar/all-
sky camera comparisons, in that it would be expected that spatial peaks
in satellite-measured electron precipitation would give rise to discrete
auroral forms that would be recorded by all-sky cameras as visual arcs.
Other investigators have not been able to well substantiate this relation,
but few believe the relationship should not exist. We shall also study
the question of whether sstellite-measured precipitation is sufficient to
provide the ionization that must exist in regions from which we obtain
radar echoes if these echoes are to be observed.
6.3.1 Correlation of Regions of Peak Energy Flux
with Locations Radar Aurora
The data from 14 satellite passes have been studied to
seek a variety of correlations. Table 6.1 presents one set of correla-
tions of these data. The first column of Table 6.1 identifies satellite
pass number.
We have categorized the type of auroral event that was
occurring in Column 2 according to the nature of the satellite precipi-
tation data. If there is no symbol, the event was a weak precipitation
event, meaning that the particle fluxes were quite low but were measurable
in some parts of the trajectory. If, in our judgement, the energy flux
of electrons (not the number flux) was large, then we have placed an E
in this cclumn. if proton precipitation was weakly observable over a
reasonable portion of the orbit, "we have placed wp, meaning moderate to
weak total proton euergy. Three events have been labeled with a P because
118
4.4
0.a a
a. .. 0 54
.0 0) VC 0 a c .0
~.va S. S o SE.i 10 4i :1V 0 0' S
4S. 00 k 0 0000 '00 0 v 130 * 40 000 0;a0 0 0 0 0 0 0 a a 0 00 c 0
a W 0 C0 0 ~S. 0 ~ol 0 fu 00 00- 4.
V V0 0C a a S. -.0 a i 136
0 ra 4. S. S.0 I 4. X 0 Ixix aCJO0 C3 34 w a
3. C. - - 4 - - - - 0 0
00
'0 S. a4 t..0 X.
.0
0 ~0 00 I-
93 0i CS. S
03 z. C-3 N 00 03 v V l r04.
0~. =~l a~.S aa aS. 0
0a ~0
0 .0 c~0 .0 O- ~ .40 0 a C3 4. 4
4.4 N N O 0S. .-4 0 S. S. 0 ~ 4
rj 4 0 M4.S Sc-: C . 1 0 -ý ' '0
O.- 00 0 04 - o Vi> - = rE- . IV. S4: 3 a C3
V' 0 01 -q L) 0 S. >4 4 IQ m2x : 4a 0. 0- ~ 0 '
4- aý13 -j
0 0 a a3
C:
0 C-VI 3 r
U
+s. ca03 0 - - -N N C C OC C- M~ 0-
0 '0 41
0 0
m V4.V t l Q s
z
u S.
0 - w mCaCIS v
119
25
the proton energy flux was very large. For our own work we have named
these "proton events."
The next column gives the College, Alaska three-hour
magnetic index for the period of the satellite pass.
In the electron-precipitation columns we have defined
collocation of electron energy peaks and radar aurora as coincidences.
This means that radar aurora seemed to be located in the same regions in
which there is a local maximum in the energy flux of precipitating elec-
trons. We have also counted the number of isolated radar aurora and peaks
in electron energy flux that were not associated with each other, and have
called these anti-coincidences. We have done the same for proton pre-
cipitation. The last column gives a measure of the simultaneity of the
radar and satellite measurements (on some passes the ephemeris data
available at the time that our data were reduced were less accurate than
we would have desired).
The determination of these correlations is extremely
subjective. In some areas there will be many, closely spaced, intense
precipitation peaks. In some cases the radar auroral boundary may be
located at the particle-precipitation peak. We have tried to be consis-
tent in calling these coincidences. The subjectivity of the interpretation
srsuld be evident by the way ue have filled out our columns.
Our attempts to redo the electron precipitation columns
always produce a slightly differing coincidence set--that is to say, our
estimation procedure is very subjective. This is very disturbing, but
we have not discovered a way to uniquely define precipitation peaks. Our
reevaluation of coincidences--though very subjective--does show a tendency
for radar aurora to be anti-coincident with peaks in electron-energy
precipitation except for the cases of very active aurora.
120
Examination of Table 6.1 as given here reveals the
following:
* In the auroras that we have not categorized as
proton events and not labeled as P, there are no
coincidences between radar returns and peaks in
proton precipitation.
* There are some coincidences between peaks in
electron-energy fluxes, but there are many more
anti-coincidences.
. For two of the proton events, since radar aurora
I is nearly everywhere, there are many peaks in
energy deposition, so that coincidences are
everywhere and therefore may not be meaningful.
* The electron-coincidence enumeration suggests to
us that coincidences between peaks of electron
precipitation with radar aurora is almost a random
phenomenon. There appears to be no clear-cut,
absolute trend, though there are more "valid"
anti-coincidences than valid coincidences.
In summary, we believe the data in Table 6.1 indicate
that except for intense proton events, radar aurora is not coincident
with peaks in proton energy flux. Peaks in electron precipitation may
or may not be coincident with radar aurora; we have not discerned a
meaningful trend, although there seems to be a tendency toward anti-
correlation.
We believe ionospheric electric fields play a role in
causing radar aurora. Electric fields are caused by convecting magnetic
fields. If the ionospheric ionization is great enough, then magnetic-
field convection is slowed in the highly conductive ionosphere. Nearby,
121
where ionospheric electron densities are lower, large electric fields
are not short-circuited by high conductivity. One might therefore expect
that at locations where particles are heavily precipitating, ionospheric
electric fields will be short-circuited and there will be no structuringof ionization into irregularities and therefore no radar scattering.
Thus, during moderate auroral activity, regions of most intense (by par-
ticle number) precipitation should not coincide with regions oi radar
echoes.
On the other hand, if magnetic-field convection becomes
extremely intense, even in regions of large precipitation (and therefore
high E-region electron densities) electric fields may reach and exceed
a threshold for production of plasma instabilities that lead to generation
of E-region irregularities.
Therefore correlation between peaks in precipitation
intensity along the satellite path and no radar aurora at the same loca-
tion might be found at times of moderate or weak auroral activity. During
strong aurora one might seek and find an inverted relationship. Table
6.1 does not support this hypothesis.
As an alternative approach to organizing data correspond-
ing to the above intuitive guidelines, the histograms of Figures 4.E.1
through 4.E.14 were made. If the above ideas concerning electric-field
4 short-circuiting in regions of high energy precipitation were valid, then
the histograms should show more energy precipitated (larger ELE values)
in regions where no radar aurora was observed. Examination of these
histograms is consistent with this hypothesis except for five passes.
Passes A518 and A641 seem inconsistent, but the time simultaneity was poor.
Passes A232, A751, and A369 are not consistent with the hypothesis.
Passes A2995 and A2979 are consistent in the sense that these passes
occurred during very active aurora.
i122
I A
i- Examination of the other histograms of Figures 4.E.1
through 4.E.14 show no other obvious consistency or trend.
An additional attempt to reveal some relationship between
precipitating electrons and the peesence or absence of radar aurora in
S~the same region of space was made by computing average quantities from
the histograms of Figures 4.E.1 through 4.E.14. Figures 6.6 and 6.7
provide summaries of averages obtained from the histograms.
The top subfigure of Figure 6.6 presents average TSUM
for each satellite pass plotted against sequence number according to
ascending three-hour magnetic K index. Open circles are the average
TSUI for the regions of the satellite pass where radar echoes were simul-
i taneously observed; solid points are for the regions of the pass whereII no radar echoes were observed. On Pass A676, which is number 7 in our
* ordered sequences of passes, no electrons were measured above counter
t threshold in regions where radar echoes were observed. Thus, no point
Sis plctted there for radar echoes. The middle subfigure presents the
same kind of data for the average electron power (ELE). The bottom sub-
figure presents the average electron energy as a function of satellite
pass according to ascending three-hour K index. In the bottom figure we
have provided a linear regression fit to the data points. The fit for
the data without radar echoes is shown as a solid line. The fit to the
points corresponding to radar echoes present is shown as a dashed line.IA
The fit was made with sequence number as abscissa; the fit serves only 3
to dramatize similarities and differences between average electron energyInin regions of radar aurora and in regions of no radar aurora.
Fig-re 6.7 plots the ratio of the various average quan-
tities of Figure 6.6 versus satellite sequence number. The top subfigure
in Figure 6.7 presents the ratic of TSUM from regions of radar aurora to
the average TSUM in regions of no radar aurora. This ratio is then
plotted versus satellite sequence lumber. The middle subfigure presents
123
this ratio for ELE. The bottom subfigure presents the ratio of averageI precipitating electron energy in radar regions to that in other precipi-
tation regions.
The top subfigure of Figure 6.7 gives the impression
that, on the average, the precipitating electron number flux is lower in
regions of radar echoes than in other regions. In fact, this i3 only
true in 8 of the 14 passes. However, the geometric mean of the 13 useabl.e
passes (Pass A676 had no electron precipitation in regions of radar
aurora) is 0.81, meaning that in expectation the average number !lux of
the precipitating electrons in radar scattering regions is 80&r of that
for electrons precipitating elsewhere.
The data from Figures 6.6 and 6.7 cencerning the energy
flux (ELE) seem to show more definitive differences between precipitation
characteristics where radar echoes wo're obtained and where they were not.
During only three passes of 14 was ELE greater in regions of radar aurora.
If this ratio was a random quantity, then the probability that three or
fewer passes as we have found would have ratios greater than 1.0 is only
2.9%. The ratio measured on two of these passes was only slightly greater
than 1.0. The geometric mean of the 13 calculable ratios is 0.55, sug-
gesting that, on the average, the energy flux is only half as great in
regioDs of radar ehcoes as it is elsewhere where precipitation inxensity
is above the satellite threshold. This averaged characteristic is con-
sistent with our often-stated hypothesis that more precipitation produces
i: .. e ioni. )n, resulting in short-circuiting of ionospheric electric
fields.
The data plotted in the bottom subfigure of Figure 6.6
show a decided precipitating electron energy trend with 3-hour magnetic
K index. The data suggest--and the corresponding subfigure of Figure
6.7 further support--an indication that on the average, electrons that
124_ _
; I _ _ _
precipitate into regions from which radar echoes arise have lower energies
. than do electrons that precipitate elsewhere. On 8 of the 13 useable
passes the average precipitating electron energy was lower in regions of
radar aurora than elsewhere. On tuo passes the average energies were the
same 4n the radar and non-radar regions. On three passes the average
Selectron energy was greater in regions of radar aurora--but the largest
ratio computed was only 1.08--which value just barely exceeded 1.0. From
these data we compute a geometrical mean energy ratio of 0.84.
6.3.2 Production of Nighttime E-Region Ionization
by Particle Precipitation
The previous section showed that regions of peak energy
deposition did not seem to be located, necessarily, in regions where
radar echoes were obtained. In this subsection we study whether any pre-
cipitation at all was occurring in the radar-reflecting regions, to de-
termine whether particle precipitation can explain the existence of
nighttime E-region ionization necessary for these reflections.
All 14 satellite passes occurred during E-region darkness.
Radar aurora requires the presence of E-region ionization. Studies of
radar/visual aurora correlation show that the visual forms do not neces-
sarily correlate well in position with radar raturns. One is therefore
led to wonder what produces the iopization that is necessary for radar
scattering. The concern here is whether there is any particle precipita-
tion at all in regions from which auroral echoes are obtained.
The particle-flux precipitation in regions from which
radar echov- -ere being obtained h.. *een investigated even if these
were not regions of peak energy flux. Table 6.2 presents these results.
On two of the passes listed in the table there were radar echoes from
regions in which particle-precipitation levels were below the detection
j threshold of the satellite. On pass A641 the simultaneity of the
125
I_- ~ ~ .- --- 3Z
S.~0 0 4J t a4J ac r= 0,.
E- -44 4'-4 &0. 0 M 0sc ý VS1JS JS ' 0= 0 4 4 0 0r
.0
C6 0 U00
CS 4 0 0 0 0 00 0 0 0 0
0
S..0
0 a 0
0 a m m w
0, 0 0 0 0 0 0 0 0 0 0 0 0 0
U 0
v
~~~ . 0 0 C02 N' fa to -40 - ) ~ - '
cz E oEa W .1 4 o t0m4
4 '
0
012
measurements was poor due to poor pre-pass ephemeris data, so that com-
parisons probably are not too meaningful. On Pass A388, 11 of 12 satel-
lite data points showed no detectable electron or proton precipitation
above our imposed thresholds where radar echoes were being obtained.
However, in this case the particle flux nearby was decaying very slowly
as the satellite approached the scattering regions. We are led to believe,
but cannot prove, that the parLicle flux was below, but very near, the
threshold detection level. On Pass A2979 the satellite data supplied to
us by Lockheed does not extend as far as our rndar echoes. Thus, lack
of displayed precipitation in the most southerly portion of the radar
echo is not of any significauce.
We estimate that, provided particle energy is great
enough to penetrate to 110-km altitude, the threshold detection flux of
the satellite counters produces at this altitude an electron density of
about 7 X 104 if the particles are electrons, and 2 X 104 if the particles
are protons. These data suggest that there was some particle precipita-
tion flux wherever radar aurora was found. The data of Table 6.2 and
our othe- inferences imply that the nighttime, widespread particle pre-
cipitation as measured by the OVl-18 during our 14 passes is adequate to
produce the ionospheric electron densities needed to explain observed
radar aurora.
6.4 Discussion of Correlation Study and Other Possible Relationships
When comparing particle precipitation with radar echoes, the
question arises whether one should compare energy flux or particle number
flux deposited. We presume but cannot prove that the energy flux is a
measure of the local strength of the auroral disturbance. From a radar
point of view the particle-number flux is a measure of the generation of
E-region ionization.
127
The particle energy spectrum is an indication of the depth of
penetration of the particles; thus it is an indication of whether parti-
cles can penetrate deep enough to produce the ionization at altitudes
C where radar aurora is observed. Furthermore, the deeper the penetration,
the lower the altitude where ionization is produced, and the greater
the Hall conductivity and the normal magnetic-field-independent conduc-
I ti?ity. As a result, the total, local ionospheric conductivity becomes
much greater, so that magnetic field convection in these regions will be
suppressed.
Conductivity is greater for large particle fluxes coupled with
deeper penetration; this may mean that high conductivity is nearly cor-
related with higher energy fluxes. We suspect, then, that the energy-
flux peaks in the satellite-pass profiles would correlate best with
magnetometer data that record location and intensity of the auroral
electrojet.
Past correlation work between location of visual and radar
echoes 2 has shown that radar echoes are often adjacent to but not super-
posed upon visual arcs. This suggests that electric fields are lower
inside the arcs because the high conductivity of the arc short-circuits
the E-field. Barium-cloud releases1 ° as well as the satellite-measured
electric field 1 1 also suggest higher electric fields near to but not in-
side visual vrcs. Thus, a correlation between energy flux and radar
echoes is a proper correlation to study. Our results indicate no dis-
cernible trend, though there may be a slight tendency to be anti-coincident.
The data of Figures 6.6 and 6.7 show that on the average the precipitation-
energy flux is lower in regions that give rise to radar echoes than in
other precipitation regions.
A radar/particle-flux precipitation correlation is meaningful
in that we need a source for the ionospheric electron density. Our
128
I__
work has indicated that widespread particle precipitation as measured by
the Lockheed satellite is adequate to produce the needed ionospheric
electron densities.
Some aspects of the data presented in this report have not been -5
studied in depth. It would be important, we believe., to digitize the
radar auroral echoes and compare absolute echo amplitude with precipita-
tion characteristics. Our estimates of particle-stopping altitudes and
resulting ionospheric electron densities, derived from work by Rees 6' 7
and Chamberlain, 9 need to be altered to agree with spectral distributions
that are appropriate for the actual satellite-measured precipitation.
Data that were recorded by Lockheed but not uade available to us concern
the shape of the electron spectrum. That spectrum is sampled on-board
the satellite at five energies. The data made available to us concerned
only the absolute flux and the mean energy. Energy-distribution details
would be of considerable value.
1N
129
~~Lj
cc 7 )
~0 - 0
25. / A1 ' 00
-J -
22 0.
22W 0 !
EC. -0.L
. 0~
LW
0 00o
cc
00
0Z l - 0
L4% 0
x Z ILo0 0 =0-L
< W
¶3 N 0 WZD alt z 130
J ~zw0 >
I a. 0
Iw cI2
0 I-- cc0o.L 0...,
OD0N Z
> <0
ILI0a 2i -m >- w
ani-UanWuP w CD
x 0- x -< 0 ILO U w k.
z cc 133
0 0o~
C
S 0=
CL3
oco
- 0~0-0
CD
3 CL~0
0.
x Z
z -
I- E.
F-
1111..- ,'"
0 a-
CM
0 1
I,
134U
U-
hil II I hi l 1II34L
>109
0 RADAR ECHOES
0 NO RADAR ECHOES
106
: log
'• !'°P i l l I •I .,,i I i i
' oRi I i I 1 1 1 1
%_
w !I
1088
io6 i I i I i I I I i I I •6 10711 ~ li i-
" o -] ! ! I I I I I l! I ! 3 HOUR -
S0 0 1 1 1 2 2 3 4 5 5 6 7 7 K-INDEX
,• '0 0 PASS
Oa t . c ' NUMBER
S1 2 3 4 5 6 7 8 9 10 II 12 13 14 SEQUENCE
NUMBER=
S~FIGURE 6.6 AVERAGE PROPERTIES OF PRECIPITATING ELECTRONS versus SATELLITE
S~PASS SEQUENCE NUMBER ARRANGED ACCORDING TO ASCENDING
• THREE-HOUR MAGNETIC K-INDEX
1135
6 0f~5
~ ~• ~ ->
SI
I,.°
110
I- C- 0.
10
to
.; W
0
0. Z4 I I I I '
10
0
.w -i
Iu0.1
10
Z V
"'o • I I I I I I I j
3UOU0.
0 0 1 1 1 2 2 3 4 5 1 6 7 7 3-HOURIn m
o to CD C.) N 0 w "r cm C9 NUMBER
1 2 3 4 5 6 7 8 9 10 11 12 13 14 SEQUERNC
FIGURE 6.7 RATIO OF THE AVERAGE PROPERTIES OF PRECIPITATING ELECTRONSIN REGIONS WHERE RADAR AURORA WAS OBSERVED, TO THOSEPROPERTIES IN REGIONS WHERE NO RADAR AURORA WAS OBSERVED,
versus SATELLITE-PASS SEQUENCE NUMBER ARRANGED ACCORDING TO •ASCENDING THREE-HOUR MA.GNETIC K-NDEX
136<
35
7. CONCLUSIONS
This section summarizes what this research effort has revealed.
Comparison of the location of peak precipitation energy and number
density of electrons as given in Figures 4.C.1 through 4.C.14, with simi-
lar quantities for protons in Figures 4.D.1 through 4.D.14 shows that,
with the exception of Passes A2964. A2995, and A2979, protons precipitate
to the south of electrons. This was the case for the eleven passes at
lower K indices without exception. For the three passes that took place
during high magnetic activity, the satellite precipitation data were not
provided far enough south to verify that this behavior was also present
during these three passes.
The data on these 14 satellite passes indicate that the electron
density in the nighttime E-region that is necessary for radar scatteringis probably produced by electron precipitation.
With the exception of the three passes that took place during very
intense magnetic activity, radar aurora was never observed to be collo-
cated with regions of peaR proton-number flux or proton-energy-flux pre-
cipitation. On the three passes that took pLace during high magnetic
activity, as we indicated above, satellite data at more southerly lati-
tudes were not available. so that it is not knonn whether the proton data
that were obtained for these three passes contain the most significant
proton precipitation. Thus, it cannot be said with certainty thal the
coincidences between proton precipitation and radar aurora on these three
passes contradicts the conclusions that have been drawn from the eleven
other passes. We believe that this anticorrelation between protov pre-
cipitation and radar aurora is significant.
137
Attempts to find some deterministic relationship between the de-
tailed characteristics of the precipitating electrons and the regions
from which radar aurora is present have not produced any unique relation-
ship. There appear to be some statistical trends, though these are
subtle. These trends are as follows:
(1) There may be a tendency for regions of locally more
intense electron-energy flux to anticorrelate with
radar aurora, though this is not an absolute tendency.
(2) There may be a slight tendency for the average particle
number flux--TSTh¶--to be less in regions of radar scattering
than in other regions where measurable electron precipi-
tation is taking place.
(3) There is a high statistical likelihood that the average
electron precipitation energy flux (ELE) is less in
regions from which radar echoes are obtained.
(4) The average energy of the precipitating electron is, on
the average, less in regions of radar aurora.
r 138!Uo I
8. SUGGESTIONS FOR FURTHER WORK
Radar data for 54 additional satellite passes have been reduced at
SRI. Comparative studies must await delivery of Lockheed particle data
which appears unlikely to occur due to Lockheed funding limitations.
Comparison between these additional 54 satellite passes would certainly
be helpful in understanding aurora phenomena, in that the statistics of
the phenomena will be greatly improved. Data from input-output experi-
ments are still very rare. J
The fourteen satellite passes that ha',e been reduced have not been
as thoroughly analyzed as is possible. The radar data that are presented
in this report were reduced for the time period corresponding to the pre-
dicted satellite ephemeris data. As a result, the simultaneity of the
radar data that we have analyzed, and the satellite pass data, is in some
cases very poor. This lack of good simultaneity can be corrected for
much of the data.
Many data correlations involving radar auroral locations, radar
auroral signal strength, precipitation-particle energies, particle fluxes,
etc., that could be performed have not received attention. In order to
do an adequate job on future correlations, the radar data must be digi-
tized and then recompared with the digitized satellite data. We beLieve
such 2 rerun atA a second analysis is mozo- important than would be addi-
tional work on the 54 satellite passes for which Lockheed has not supplied
us with precipitation data. For iur comparison work, the data that
In order to obtain the Lockheed data for 14 passes analyzed here it was
necessary for SRI to fund Lockheed directly.
139
Lockheed delivers to us in the future should give, if possible, the in-
dividual electron channeltron counter rates rather than the summary value
EMN. This would help us more accurately determine from the satellite
data the altitudes at which the electron-precipitation energy is converted
to ambient E-region ionization. For our comparison work we need to extend
the work of Rees 6 ' 7 and Chamberlain9 to better meet the needs of this
progran. In particular, we need to compute depositioil for particle spec-
tra and pitch-angle distributions that are more 1'., those that the
satellites actually measure. Furthermore, we must look into the effects
of proton-charge-exchange scattering in order to do 3 better job of making
spatial comparisons between aurora and proton precipitation.
140
I!_
REFERENCES
1. R. L. Leadabrand and J. C. Hodges, "Correlation of Radar Echoes
From the Aurora with Satellite-Measures Auroral Particle Precipita-tion," J. Geophys. Res., Vol. 72, No. 21, pp. 5,411-5317 (1 November
1967).
2. W. G. Chesnut, J. C. Hodges, and R. L. Leadabrand, "Auroral Back-scatter Wavelength Dependence Studies," Final Report, Contract AF30 (602)-37345 SRI Project 5535, Stanford Research Institute, MenloPark, California (June 1968).
3. T. H~agfors, R. G. Johnson, and R. A. Powers. "Simultaneous Observa-tion of Proton Precipitation and Auroral Radar Echoes," J. Geophys.Res. Vol. 76, No. 25, pp. 6093-6098 (1 September 1971).
4. J.E. Evans and R. G. Johnsoa, "Auroral Input-Output Measurements,"
Contract Nonr-3398(00), Task No. NR 087-135, Lockheed Missiles andSpace Company, Palo Alto, California.
S 5. G. B. Shook. "A Coordinated Experiment on Radar Clutter and AuroralParticle Fluxes," Final Report, Space Sciences Laboratory, Contract
N00014-69-C-0207. Task No. NR 087-170, Lockheed Missiles and Space
Company, Palo Alto, California (30 April 1970).
6. M. Rees. "Note on the Penetration of Energetic Electrons into the
Earth's Atmosphere," Planet Space Sci., Vol. 12, pp. 722-725 (1964).
7. M. Rees "Auroral Ionization and Excitation by Incident EnergeticElectrons," Planet Space Sci., Vol. 11, pp. 1209-1218 (1963).
8. I. G. Poppoff and R. C. Whitten. DASA Reaction Rate Handbook, Chapt.8, "Reaction Rate Data From Atmospheric Measurements," DASA-1948
(October 1967).
9. J. W. Chamberlain, Physics of the Aurora and Airglow (AcademicPress, New York, N. Y.. 1961).
10. E. M. Wescott, J. D. Stolarik, and J. P. Heppner, "Electric Fieldsin the Vicinity of Auroral Forms from Motions of Barium Vapor
Releases, J. Geophys. Res.. Vol. 74, No. 14, pp. 3469-3487 (July1, 1969).
141